Webinar: What’s the Future of AI in Marketing?

Machine generated transcript…

Hello, everybody thanks for joining were just going to give it two minutes just for anybody coming back from lunch, just just to join the call and well be cooking off in a minute or two okay, so hello, everyone and thank you for joining our webinar today on Whats, the future of AI and marketing. My name is Sherman Mullen, and it really is my absolute pleasure to be joined today by Jeremy white for those of you that dont

know its Jeremy let me embarrass him and give it a quick intro Jeremy is a chief strategy officer here at whats marketing he has traveled all over the world delivering presentations and speaking to marketing execs about how whats an AI revolutionized the future he is the host of a popular ten words podcast and has also authored three books the most recent of which is also called ten words but before I pass out Jeremy I just wanna go to some quick housekeeping tips for everybody we will.

Have a live Q & A after the webinar. So. Throughout the webinar if you have a question please type into the box on the left hand side. Of your screen and we will answer them to five one-year once the webinar is finished and trust me. I still have a slide essay and this is going to be a great webinar and one you’re going to going to want to refer back to and aid you in doing. This were going to leave the webinar up for download afterwards and to access that

its going to be the exact same link they used to access the webinar so please feel free download it use of for reference in the coming months and years okay so that’s more than enough for me now over to you Jeremy cheers thank you very much Sean thank you for that very kind introduction as well hey everybody how you doing good morning this is a lot more people than I thought we would get.

on this is fantastic we have people from all over the world as well so good morning good afternoon or evening depend upon whereabouts you are let me tell you where this came from before we got started I spend a lot of my time talking about what do we do and why do we do it and why I should anybody care and I think there was even a playful title that we put together for this which is a WTF AIA marketing which of course means whats the future of AI in marketing but.

there’s also so many myths around AI is going to be festival of marketing next week which for those you in the UK is one of the biggest marketing conferences that we have and youve now got literally the world and his wife has attached the word AI to just about everything and some of that is true some of its not true some of its just blatantly false and BS and so.

what I thought would be really useful is first of all just to explain what AI is we have the biggest day i platform in a way Im not insect ills I just want to try and give you an open view of what weve got and how it works and kind of what our view is as IBM but also just to try and put a spotlight on some of the places that marketers go wrong because wed love to get seduced by shiny things.

were going to look at that in just a second so I really wanted to try and look at all the myths and stuff behind the metrics as well so look at AI what it is what it means what the implications are going forward I have visibility of quite a lot of research as well in my roles so Ive with a very fast-paced quite a visual deck some of you’re going to be multitasking while this goes on Ill encourage you maybe to pick on a few slides you might want to screenshot them you can see the deck.

afterwards but its really just to do a little bit of a brain dump about what I see about AI with the biggest challenges and where its going to go so that its basically preparing you guys to try and be better at your jobs going forward and there’s a couple of reasons for this and the main reason I guess to start with is we have this very strong belief we like I said we do a huge amount on research we have a whole research division we put about six billion dollars a year into.

research as a whole we register about 9,000 parts a year and as it goes about half of those are in AI in cloud quantum computing so we invent a lot of very very cool stuff with a lot of very smart people someone tell me were actually have six Nobel Prize winners as part of our our science and research team and a lot of the stuff that they get to in been people dont realize what it is and were going to try and show some of that.

today so were going to have some some old stories Im going to show you some very new stories Im going to share with you a case study which has never seen the light of day which is super super exciting which is going to speak very much to where the future of marketing is going but the only reason I say those stats is not to kind of chess beat iBM is amazing which obviously of course we are but hes just.

to say that we survey a lot of people just to figure out the challenges that going in across the industry and our insights are showing that on average about four out of five executives this has probably been true for about two and a half years now four of five executives 80% of executives feel overwhelmed and under prepared for the challenge just their business is going to face over the next five years now what Im going to try and do throughout this webinar is not just try and give you a few insights into AI but Im also going to make a.

case that the next five years are probably going to change more than the last twenty now that’s quite a big statement there’s the headline right there isnt that you know weve gone through a lot of disruption weve just had the ten year anniversary of blamin going down you know weve just had the 20th anniversary of Garry Kasparov losing to IB I mean that famous deep.

blue match the chess game a while ago will cover all of that in just a second but its really just to try and set the scene so that you guys know where it is that were going to be heading over the next hour because a lot of this how can I put this a lot of this feels like were talking about innovation it feels like were just talking about all these shiny new technologies that do wonderful things that help you to do your job faster or helps you to run your business faster or to compete faster maybe you’re.

a smoker only trying to compete with a big cup and just like big social media did years ago ai is now allowing allowing the playing field to be flattened so the smaller companies can use this really transformative technology to compete with the big companies some people mistakenly call that innovation but I just wanted to point out at the top that all the innovation is is doing the same thing as everybody else a little bit faster this is a great line I heard from my friend brian solis one of the best analysts i.

think in the marketing industry and what brought only went on to say is that actually innovation and disruption and a lot of marketers get this mixed up because if innovation is just doing me the old thing faster than everybody else then disruption is doing a new thing that makes the old thing obsolete completely different point of view isnt it a new thing that makes the old thing obsolete Bill Gates I a couple of years ago said that whole thing that if you aint broke dont fix.

it doesnt apply anymore because what he says is if hes saying if it aint broke its obsolete you know that’s read the original move fast and break things mentality came from it Facebook now that’s got them into a little bit of trouble over the years and were looking at some type of stable infrastructure which is where Facebook of now pivoted to but really everyone in the tech industry IBM included is looking at how to disrupt.

markets and if there’s one technology most likely to disrupt over the next couple years its certainly gonna be a eyeing though well also look at things like quantum in just a second because that’s where the worlds going to go after the next frog yes but let me just start with this really this this idea that maybe you’re not expecting and well get into some of the stats in a day – in just a second Im going to set the scene and just show you some of the.

dates that might knock your socks off absolutely up-to-the-minute 2018 going into 2019 stats about the state of the marketing industry but before we do all of that let me just share with you one of my favorite people in the world this is Steven B Johnson so Steven with a V this is some amazing TED Talks anthropologist really great philosopher did a fantastic TED talk that Chris Anderson actually created into this.

little eight-minute slot and built kind of paper cut out animated video youve absolutely got to go and check it out because what Steven B Johnson said in the middle of this TED talk he said look if you want to see where the future is going you want to look where people are having the most fun but we dont play enough and in fact Chris dicks and whos.

the block chain VC that heads up the big Silicon Valley VC company an increase in the Horowitz Chris Dixon said in a blog post that went viral a couple years ago he said if you live work hes going to look like in 10 years time just look at what really smart people are doing for fun at the weekends so I only wanted to.

set that up because a lot of this technology which Im about to talk to you about has come out of play has come out of our researchers and our scientists just playing with cool stuff to see what is possible and sometimes a commercial intent comes afterwards and that’s important to remember because I think some marketers forgot what its like to have fun were so deep in the technology and the analysis and reports and the data and architecture you know my tech seems to have taken over sometimes we forget about the creative side of going away.

and having fun and Steven B Johnson honestly check out his top he goes through the history of Technology going back thousands of years and says every single moment in time when something cool has been invented that went on to revolutionize the world something truly disruptive it always stopped at it from someone playing with technology so just have that in the back of your mind as we go through these slides because that’s really what I want to encourage you to do because as we go through some of this stuff what are you going to see very.

quickly is that the case that is Im talking about our brand-new incredibly cutting-edge emerging technologies and a lot of them dont even exist in marketing at the moment that’s why Im so proud not just to be given this type this this talk but because you guys are here listening because youve obviously got an appetite and a hunger to try and have that competitive edge and there’s a lovely line from Seth Godin he once said.

one of my favorite marketing authors Seth Godin said if you wait until there’s a case study in your industry you waited too long that’s a great line isnt it and that’s really some of these technology because when you look at something like the Gartner hype cycle and Im just taking this as a shot this this comes out once a year its a great report you should absolutely get your paws on it if you can find it speak to your garden a rare poll and show your companies have got a handle on a lot of.

this has been shared on Twitter a whole bunch of times really all this is showing is the shiny technologies that marketers get excited about and we love getting excited about showing to show any things and basically what this means all this chart is showing you is called the law of diffusion of innovation for those of you familiar with you know simon Sinek and that law of hype technologies coming in you know and then.

at what point they mature and become mainstream and all this chart is really saying is that when something becomes mainstream arguably that’s when 40% of the industry has adopted it but in that innovation trigger that little column on the left hand side that’s when less than 3% of the industry is adopted so these are all the topics that we see.

floating around in wired in Fast Company when you go to conferences sure itll be a festival of marketing next week just like use it to Mexico last month or South by Southwest or chaos summit all of these big tent neurology conferences Dreamforce was last week people talk about all the things and a left hand side whereas its often of things than the right hand side are the things that.

have fallen off the graph like email marketing and personalization that are so mature they show economic value and Im just going to pick on one in just a second but the quote that I like to sum this up it actually came from the guy that invented Wired magazine himself Kevin Kelly no Kevin Kelly wrote an amazing book called the inevitable the twelve tech force is that going to shake the future you should absolutely go and read that put it into language that every guy on the street can understand about the forces are going to shake the.

next few years and and what Kevin Kelley says is that there’s no business case for innovation you should perfect what you know first and that’s to say that sometimes you get excited about the new thing coming through I was at a conference last week with some very senior CMOS blockchain for advertising was absolutely dominating the conversation how blockchain is going to turn the whole.

advertising supply chain upside down but yet the majority of the businesses in that room werent even sharing their own customer data between their own sales marketing and customer service teams so that’s what Kevin Kelley says when there’s no business case for innovation hes like make sense of the data that youve got before you start investing in some of these sexy of things on the left-hand side now of course its exciting that’s why you guys are on this webinar today but what youll see that came on the chart for the first time last year is AI or marketing and youll.

also see then its got a little blue dot next to it which is really what that’s saying is that three percent of the industry is currently looking at AI and its gonna at some point mature into the business strategy of an organization over the next five to ten years now now lets take some Q&A afterwards about whether or not you guys think that’s true and by all means grill me with as many teddy complicated future gazing questions as you can about such strategy and the reality where they should I should live.

within your business strategy because I whilst I know that it is a very hyped technology and it is very early days I certainly dont think its three percent its going to mature over five to ten years in fact with it consultants think last year I did some independent research nobody knew it was IBM in the time in that commission marketers across the UK in Western Europe to say okay whos putting AI into their marketing strategy and over what period of time are you looking at implementing this and.

what I found that was quite different I found out that it was 28% edenia marketers were going to put AI into their strategy at some point over the next 18 months to two years now that tells you something interesting that tells you that there’s still a window of opportunity where nobody else is looking at some of this technology because they.

havent figured out what to do with it yet but it also shows that the rates of adoption is much faster than people expect so that’s really the main takeaway from this but since were kicking things off about AI it would be remiss of me not to give you the official explanation according to what analysts like Gartner think it is so here we go artificial intelligence for marketing AI for marketing comprises of systems that change behaviors without being explicitly programmed based upon.

data collected usage analysis and other observations for marketing use cases this enables technologies and techniques then this is going to include machine learning deep learning and natural language processing now this could be anything from an AI looking at vast amounts of data I was with a large Im going to say Fitness company this morning theyre looking at an AI that can figure out the hashtag of when.

somebody says something and the AI can then search every single post that they ever post and then try and see the mood and the emotional intent and the personality of that individual based upon every post that they’ve done but the initial trigger is then using a hashtag of this specific product and then you can start looking audiences and moods and behaviors and predicting whats going to next now there’s something interesting in that as a point because when I say.

predicting what were finding at the moment and Im not going to name any names but you may know who they are there are some big markets in clouds out there claiming that they’ve got an AI a lot of the time all it is is a very clever predictive engine or maybe even prescriptive its not showing whats going to happen next its maybe giving you an insight the data says this we think this is going to happen next and because of that you should do that and Ill thought of it is reals based.

and there’s some element of machine learning maybe but that’s not a I will look at exactly what AI is in just a minute but its very very careful that you distinguish between AI and predictive because AI is really looking at unstructured data massive massive amounts of data think about it 90% of all the data in the world was created in the last 12 months that’s insane right and its going to double over the next.

18 months almost in line with Moores law about 66 percent year-on-year so the fact that datas doubling every 18 months 90% of it was created in the last year and were looking at all the data that businesses sat on about 80% of it is unstructured its video its voice its text its gonna be things from social media from IOT from connected.

products from a million different places this isnt the structured data like things that live within spreadsheets things that are tagged properly that live with him at your database your CRM that structured data that for example is what Salesforce might do with something like Einstein what a true AI is going to do like Watson that well look at in a second whos making sense of the vast amounts of data there’s unstructured now the key which were going to look at in a second as well which is not getting too seduced by.

innovation in AI there was a line that got me used when I was listening to webinar the other day they said heres the best piece of advice we could give to any marketer focus near-term ai initiatives on data centric time intensive marketing challenges so that’s the big win there’s a lot of amazing things I was going to do and Im going to knock your socks off in a meeting with some of those use cases but the best thing that you could do is just.

looking at how you process large amounts of information very very quickly and something like Watson very open architecture Im going to show you how he could access him for free and play with him later on but hes able to process about 10 million records a second so he makes that process of automating data manipulation analysis very very quick and again that’s a competitive analysis that you can get from more companies competing with big companies so that’s setting the scene now were going to have a little bit of a whirlwind trip over the next five or.

ten minutes just through where we are at the moment and why this is important so hold on to your hats if you want to screenshot some of these stats I dont if you’re doing good stuff at the same time you might want to steal some of these slides and some of these numbers as we go through them because this is going to be a bit quick lets start off with looking at things from a brand.

perspective three-quarters of the worlds top brands are going to lose one-fifth of their brand equity by 2020 if you’re not a CMO on this line right now and I can see a few of you or if you’re not a VP direct to head our vulgar aspirations to be a CMO if you’re not terrified especially if youve not got a b2c company if you’re a brand that has a big consumer face the fact that brands are going to lose 20% of their brand equity so this is a sort of stuff.

that’s measured by inter brand their new into brand global 100 report comes out in next week I think it is you’re gonna lose that within the next two years a fifth that’s pretty scary stuff and heres the reasons why social media polarization social media doesnt work in a way it used to well look at that in just a second inflammatory content I dont need to tell you what that is ID.

blocking I dont need to tell you what that is lots of people switching that off and were going to give you some data for that in a second corruption in the ad supply chain what about that what about those three second video views that live on a tab 5 tabs deep that the customers not even seen but Yeti tracks is a view and then you have this consolidation of power when you have a few big players owning the majority of the pie Facebook and Google in digital app about 70% of all the traffic and you.

have CMOS or some very big companies quite uncomfortable at the moment because they think that Google and Facebook may know more about their customers than they do so a shift in the marketplace for sure what else have we got what about the data fifty-five percent if consumers no longer want to provide any any any personal data to marketers that’s pretty scary stuff that’s quite spicy I heard that start last year Ive checked it out again it still seems to be current he markets have backed that up and you probably know that this is.

true from your own examples its getting harder and harder to get peoples data weve got to create a compelling experience first based upon anonymous data sometimes emotions sometimes personalities understanding the mood of an audience is super super important we cant just capture peoples day two anymore that’s what we used to do we used to get people to sign in with Twitter or Facebook Connect Ill give me a mobile email date of birth whatever customers dont want that anymore weve got to provide different experiences Gartner actually calls that personification not personalization right message to the right person at the.

right time when you dont have their information that’s one of the key areas where AI can help because it can understand emotional in 10 days upon anonymous data huge huge play cant over emphasize that enough b2c companies especially the more emotional something is at Facebook we used to call them passion pillars music TV fashion film sports up to 90% of conversation and now in messaging apps and no longer and those public forums maybe you can scrape the comments and Instagram like I just.

mentioned and whether you’re Pinterest you know god forbid you might when you squared the comments and YouTube and see how the car-crash of conversations are going on their public Facebook etc etc a lot of that stuff just doesnt take place in public anymore probably messing you messaging that facebook Messenger snapchat we chat line Facebook message or Instagram DMS you know people even send in URLs on emails so that changed a lot over the last two years so what we do is we find ourself in this place.

where marketing in many cases is now the least trusted profession in the boardroom but what about that when you find that not only are less marketers than ever before on executive boards but eighty percent of CEOs according to harvest say they dont trust their see him I was anymore in fact a recent survey so a couple of weeks ago 67,000 executives and less than 2% were marketing marketing has a bad rep in the boardroom and we need to get that back and we need to get back that credibility by showing true.

economic value on what it is that we do and stop measuring some of the BS metrics that are very fluffy around reach and impressions and all that kind of stuff that we know doesnt mean anything in to people outside of the marketing industry if I was you Id want to impress the CTO and CFO more than the CMO that’s how you’re going to succeed in business and certainly with your.

marketing efforts and just about this heres the last one because it sets the scene for some of the IBM stuff Im going to show you in just a second told you this is going to be faster than I hope you still with me I hope this is all ok the average employee wastes about one day of their week searching for information that helps them to do their jobs 19.

8% of time that’s again where AI can help you have the worlds best assistant by your side that knows everything youve ever said you ask him a question I say hey hey Im referring to Watson in my case selfishly in cases that weve seen every time you ask him a question it saved you for hours and has given you a response back in 15 seconds you think what that’s going to do for the future of agile of cutting back the traction.

between agencies and analysts and partners and went backwards and far was trying to run some of these reports now again quick fly-by-night of where the industry is because in 2018 hes assumed that there’s about a hundred AI startups having the majority of the share of voice in the press now this was at the beginning of the year there was 100 I.

actually checked about six weeks ago and hes sixteen hundred already gone from what 100 to 1,600 think about that think about where Scott bring his landscape came from I know hes sick to death of seeing this Ive shown this a bunch of times but the reason I love it is because seven years ago Scott said there was a hundred technology companies trying to help you connect with your customers faster and then he started to measure them knowing what the industry.

was going to do and he just got bigger and bigger didnt it went from a hundred and two thousand eleven to three hundred and fifty if anyone in 2014 947 eighteen hundred and seventy six three thousand eight hundred and seventy six five thousand three hundred last year about 4,800 unique now whats interesting of those logos half the startups and half have got less than.

five staff that creates its own set of problems and this year were even over that six thousand eight hundred and twenty nine companies claim I think that they do almost the same thing and when you look at the company that I work for IBM and you look at the and weve got especially AI and Im going to show you some examples of that in just a.

second this is the only sale zip it promise this is the only bit where Im going to even sound a tiny bit sales it with the only company in a world that has twenty eight of forty nine categories Adobe Salesforce Oracle SAS s ap even though the small point in solutions your pegas and hubs bus dont even come close twenty-eight of forty nine categories we have a footprint in this the.

misconception is that were expensive to deal with we can get into that later on for the small use cases that is absolutely not true so as you go through your job and your career and you’re listening to the news and you’re looking at the hype cycles and you’re going to conferences just think back to this chart over the next few years because where we are at the moment its just.

going to look like absolute chaos so for the next few years so there’s never been an easier time to do your job even though it does feel very overwhelming with where we are at the moment so you need to keep up and you really really need to invest in this stuff and some of your time looking at some of those emerging technologies from more.

importantly understanding specific use cases so let me give you a few little examples like I said 20 years ago just over 21 years ago IBM beat Garry Kasparov who was a nun beat and Grand Master he won the first game we built an AI computer was really just a very predictive fast he called it a ten million dollar alarm clock which is quite nice but he also said that IBM.

destroyed his life when we won what was nice when he wrote his book smart thinking hes they said 20 in 1997 was really the year of intelligent machines this is where machines got smart and this was really the origins of where AI commercial AI started to take off now this was over 20 years ago now whats nice is that hes come full circle and I was chatting to him four months ago and he said really the best view is a good human and a machine is the best combination not one or the.

other now also you remember at the top I said when you’re looking for fun you know you want to see where the future is going look where people have it in the most fun we built this for fun for a chess game we built either fun or a game show we won jeopardy natural language processing a very complicated game show where you given the answer youve got to.

figure out the question and the natural language processing that Watson did we started building him in 2006 matured in a way we won the game show in 2011 now this is something fascinating on a point you towards because just a couple of months ago a story that’s even bigger than Jeopardy is the AI engine within Watson that was put head-to-head with the worlds two top debaters so you can see there the tune-up debate as you can.

see Project debater in the middle just for fun when you finish this go away on YouTube go in search Project debate at IBM and what you’re going to see is a very very articulate watson using language and taking from a corpus of tons and tons of information to have a real-time debate that there’s no machine learning this is just pulling from millions and millions of articles across the web across Wikipedia from social media from looking at all the headlines.

going back years and years of the New York Times in order to have debates around the future of Education should we subsidize Space expiration for example and what someone and you youve got a computer this isnt biased youve got a computer that has to make a case against these debaters to try and convince people in the audience to change their minds this really really was you know something that was absolutely revolutionary at the time so project debates are hopefully there what you can.

see on your screen let me just make sure that the webinar is working here looks like you’re still seeing jeopardy and push that to the audience okay somebody mentioned in the webcam if you can see this okay here but hopefully what you can see there we go is jeopardy okay now all these thing on your screen is the.

next generation of this this is summit and again just have a look just to try and see where the world is going because what some it is is the worlds biggest supercomputer its a cloud-based computer and this is a real-time cloud-based computer that can do in one day what it would take a human thirteen thousand years to do that’s just crazy.

isnt it thirteen thousand years and then whats happened is we built this again we built it for fun weve built it to try and stop up things like the amount of time that’s spent on clinical research trials you know you look at things like Memorial sloan-kettering we built whats in 2011 its straight away we put it into oncology going to help physicians to try and make smarter decisions about what to do and again for fun what we did is put it straight into Wimbledon these slides are changing.

quite fast so I hope you guys are with me I hope you can see this now Wimbledon is a great example and let me tell you why because this is just another use case of AI and then Im going to show you were going to put some meat on the bones in just a second because Wimbledon has 70 million fans and those fans are only loyal for about three weeks of the year now the interesting thing about.

that is that its very very important that Wimbledon is able to maximize customer engagement in a very very small amount of time with a phenomenal amount of information in order to try and provide more compelling experiences again without asking for peoples data so what they did is they came to IBM in order to try and solve some of the problems and automate some of the stuff that they got bogged down in doing and one of the things that they did was the highlights now if you go on Wimbledon.

comm if you look on YouTube if you look at any of the highlight packages for Wimbledon youll see something interesting in 2018 and Im going to show you here a picture of what this looks like we dont often show the screen very often this is cognitive highlights because 2018 was the first year where an AI did 100% of the highlights with no human involvement whatsoever we went back to 1877 with all the date when Tom Watson tennis we went.

back to the beginning of our partnership with Wimbledon in 1990 53 million points and then he learned how to play the game he learnt how to watch there was video tagging and there was voice tagging you can see excitement levels of the audience and what created the highlight packages completely on his own so it freed up the Wimbledon team to go off and be more spontaneous with social media posts with creating content of something that happened in the crowd and it completely changed the way they work because.

theyre going through a transformation Wimbledon isnt really a tennis championships Wimbledon wants to complete on broadcasting they want to be a media producer they want to compete with the BBC in the ESPN and sky and theyre using AI to do that to speed up those things that could previously have been done by a team but that’s all automated now that team goes off to do better stuff and there’s a great example that can show you here if you see the FIFA.

World Cup highlight machine again this is a use case of AI just pointing you at different examples at the moment you can go and google them afterwards this was phenomenal think about the World Cup 3 4 billion people watch the World Cup is about 46 percent of the worlds population Fox Sports was looking at that data and realizing that as those 3.

4 billion people most of them only needed to watch one minute of game footage for them to be tracked as a viewer so when you’re looking at Fox Sports in the US they tried to grow its base its looking at 280 million people online and hes trying to get them to engage for as long as possible so what we did is we looked.

at the FIFA World Cup highlight machine and we fed in every single second a video footage going back to 1958 and just like we did with Wimbledon now what youve got is that fans can go and they can ask any team any player any match any play type fouls red cards goals and Watson based upon a million hours of archived footage can create personalized highlight package which then got shared on social media and as you imagine the viewing figures and engagement levels shot through the roof now the reason I want to share that with you because.

heres the deck later this is why some of this stuff is important and like I said Im going to show you where you can go to get this in just a minute early adopter brands according to Gartner showing that brands that redesign their websites now that doesnt have to be a start-up could be an existing brand that we optimizing their site at the moment but you react to measure your site for visual and voice search there.

hes hardcore digital revenue by 30% by 2021 now when you think that by 2022 50% of our browsing behaviors going to happen without a screen things like voice technology on visual tagging technology these are all the apis that live within things like whats an AI that can help you to do that that’s going to be the basis of these brands that are able to compete and disrupt now.

the reason that this is important as well Im looking at a report right now from rich fat rich relevance and its showing that across the UK its looking at UK France and Germany only 25% of the industry aware that some of this technology exists and in fact what it did is it when the survey marketers to say do you think this is creepy or do you think this is cool in fact you can google it the reports called creepy are cool and what it showed is that the top.

five creepy technologies according to consumers are targeting with ads on your phone based upon your location to shops field fencing voice assistance within your home listening and making recommendations and emotion detection technology adapting your shopping experience based upon your mood now these are all the technologies that are completely open based upon anonymous data but at the moment people are kind of freaked out a little bit and they dont know how to make sense of all of.

that stuff so this is our challenge as marketers not just to understand what is possible but just because you can measure things right doesnt mean that you should well think about that in the Q&A maybe in about 10 minutes because just because some of this stuff is available weve also got to think about the morals and the ethics of how you use this and what the intent of your audience says dont people are the early adopters like the run the hype cycle but.

some of the laggards some people wouldnt be comfortable knowing you could understand their emotions based upon their posts or their images you can read 500 things from a photograph they wouldnt be comfortable you serving them a message based upon their mood not upon what they’ve said so if you want to have a look at a case study again I told you I was going to dump a whole load of stuff on you if you want to look at a.

case study where this is gone just go into IBM calm music and what you’re going to see is a guy that Im working with at the moment absolute Rockstar called Alex ticket he is the producer of Eminem of Rihanna of imagine dragons Nicki Minaj what we did last year was helped to build the worlds first AI powered pop song now that was fascinating because we fed in about 25 years worth of Billboard 100 we fed in the last five years of news.

articles and updated up scores from Wikipedia and then we applied all of Alex the kids social audience were huge to understand the mood of the audience in accordance there was news articles and based upon that we helped him to write a song so its a huge use case of how AI could be used to understand what people are saying on sociable and to trigger some type of event or a piece of content now he went one stage further this year this is a brand new case study.

it was shared this once before what we did this year is we found Alex was wasting so much of his time trying to find new artists and he was working with her HDR you can see on the screen that he wanted to try and figure out if AI could help him to find an artist so this is where AI comes into its own right find a meaningful signal.

in a noisy environment huge amounts of data gained find somebody that’s unsigned so I want a female artist from America that uploads one song a month shes got a tournament at downloads and she sings about certain things so using the music intelligence API within Spotify and you can see the videos and YouTube we went to search through and a half million artists we looked on 1,600 music genre genres listen to those songs listen to the tone the sentiment personality insights and based upon that we were able to find an artist that Alex.

the kid could then go off and sign and she did the collaboration the backing vocals for her now just think about that for a second think about that for the future of influencer marketing of finding the key people that are really talking about your brand of looking at strategic partnerships how is with MLB NFL MBA last week looking at sponsorships think about the implications for that Formula one are.

you looking at who to work with what collaborators its exactly the same engine exactly the same engine so as we start to bring this to a rap in the next minutes let me just give you a couple of the technical bits and again this is a page that you might want a screen shot dont have a look at this this article was called artificial intelligence for the real world and I really like it because just like kevin kelly said dont.

start with moonshots you know there’s no business case for innovation and perfect what you know first dont always invest in a shiny stuff if your customer service your markets and your sales isnt joined up properly most businesses dont do that but make sure the AI in some small cake is and Ill show you where to start just in a sec in some small case make sure that you’re at least starting somewhere but the important thing for this webinar is to understand that when we say AI could actually mean a couple of different.

things because youve got three different types of a Im not going to read off the bullet points to you now but you see the process automation or cognitive insights or cognitive engagement really what that’s doing is its looking at different types of AI some of that may be robotics and automation stuff could help with in my kissing for CRM but making you smarter is often one of the best things for AI dedicating detector and patents and large volumes of data cognitive engagements when I work with and whats the customer engagement side forcing.

clues things like chat BOTS in those product recommendations even if theyre based upon visual or voice so let me just show you a quick example of that in just a second and again I just wanted to give you some of the technical side and this is independent again you can see just to get a little bit of knowledge trying to inform and inspire right some.

of the most common examples of AI accordance IDC again that’s the definition there you looking at deep learning machine learning and conversational systems Im trying to squish about 40 hours of content here into up 45 minutes but dont have a look at that now of course Im very proud of this because it shows that iBM is miles ahead of anybody else with the only leader in AI in marketing at the moment probably about 18 months ahead of anybody else in the.

industry hopefully it stays that way well see where that goes but if you to look at this platform itself I mean I said Ill leave you with a few examples its actually really simple you can go on to IBM com whats and right now you can scroll halfway down the page and click on build with our developer tools.

what you’re gonna see is loads of apis all the free demos SD cake kits for you to mess with within your team and youll see how we process unstructured data whether its voice whether its visual search like Alex decayed and Wimbledon whether its making sense its a bunch structured data and then mapping that to CRM or speech to text whatever it is for something to be truly AI not predictive not prescriptive true AI it has to do those four things youve got to understand reason learn and interact so when I start thinking about all that.

unstructured data and you start thinking about where that data is coming from youve got the open API system this is the part of the business that I work in weve got marketing commerce and supply chain and really its just taking that data together this is the unstructured data that you guys have got to make sense of and you cant because you cant keep up because its doubling in time every 18 months its like the four PS of marketing you know really what we should be looking at now is putting that down.

into the people places and products and all of those different sources of information can be processed by an engine like Watson so Im not going to speak to this because it would feel like Im trying to sell you a product but that’s basically what Watson Marketing is with all the different elements of how it calls any piece of data from any source from any vendor anywhere in the.

world and sticks all into one place that’s basically what we do if you want to screenshot this these are the nine use cases of how marketing could help with AI so this is true AI in marketing and you can have a look at those you can google those and see afterwards couple of quick examples you could look at the North Face comm slice XPS and youll see the worlds first AI powered website.

doesnt ask you for any information its about two years old now makes recommendations that have been automatically tagged from someone uploading 75,000 products at the north face side Watsons tagged them all making recommendations based upon mood and the weather net supportable you know hackathon the other day you can upload your favorite celebrity photograph tell you exactly what theyre wearing and where you can go and buy it from that’s called cognitive fashion there’s some interesting stuff that were doing around that though last three or four slides I hope this is.

interesting I told you it was a little bit of a whirlwind if I could bring all of this together and Im not going to do the demo now what we could talk about it because I want to leave like I said 15 minutes for Q&A we have something called the Watson marketing assistant now when I go on stage or Im speaking to CMOS in private boardrooms this is usually the thing that I show because traditional marketing would take about 30 days to do what I would do using my voice in 15.

minutes and I can build a campaign you can see in the bottom corner there hundred and forty three thousand this particular demo is actually fixing a campaign with 62 million records in email looking at the subject line and the emotions that people feel when they look at that natural language and with fixing the campaign based upon my voice with people asking whats and questions about why this email campaign not work now normally you go back and forth analysts and agencies to be able to do this and this is me with no technical.

knowledge asking whats and to try and help fix it the reason this is important friends on the line is that IDC is recommending that people need to be looking at this now because within 10 years within less than a decade 66 percent of all business apps will be built by people with no coding experience whatsoever using apis like this so its so important to understand it now exactly what it is and how it can help of course.

there’s lots of other alternatives out there what some students are pretty interesting stuff are you welcome to go and check him out but let me leave you with this as a thought weve talked about AI the best way to think about this is not even augmented intelligence that’s the language that IBM users using engines like this to augment human behavior what you should be doing is thinking about it like an intelligent.

assistant this is not something that’s going to destroy jobs take jobs away all that kind of stuff this is an intelligent assistant that helps you do your job faster why because you waste about 90% of our time searching for stuff whether that’s reports and dashboards or metrics or like Alex the kids are searching for an hour so we start with garry kasparov so lets.

finish with him because this isnt about whats and taking over this isnt about jobs being replaced this isnt about intelligent machines replacing content teams at Wimbledon this is about good humans working with machines to provide compelling insights for customers because of all those stats that we shared about the landscape changing pretty quick so the technology of course its wonderful of course its amazing.

but whats important is that we have faith in people they couldnt as smart if we give them the right tools like some of the ones that youve seen today youll hopefully do something truly wonderful and magical with them so thank you very much thank you for your time weve gone through 45 minutes of super fast quickfire content the bad new all still with me and you all okay let me hand back to Sean well.

see if were taking questions from people and feel free to drop something in the chat and Ill answer any questions on anything you like something else Kim Dover to fast you wanna know more about let me know okay perfect thank you very much Jerry so we do have quite a few questions so the very first one is is this something I can use I run a small marketing agency for my small clients they might not generally have.

more data than Google Analytics plus possibly a small customer base mm-hmm do you know where I would start with that and I was with one of our agencies this morning actually so that’s a great question because it was with our respect to my friend who I was with a very very small agency in limited budget limited days as well there’s something called personality insight so when I pointed you towards at page ibm.

com slash Watson you look at the developers you scroll down youll see personality insights as a tool in fact if you search back my tweets youll probably see a link that Ive shared what personality insight does it understands the mood of your audience it could scrape Twitter everything has been publicly quoted what you can also do is you can put in.

websites you could put any amounts of raw text or any white paper and here the reason why its really interesting why it is so good especially for small businesses because it only needs 6,000 words to be accurate to understand the mood of an audience how do they respond to this campaign or this message or this copy that Im thinking even a speech that youve written Ive used this a.

bunch of times the speech is to understand how an audience might get it 6,000 words is about one hour of talking you know its a couple of white papers but you can just copy and paste that in and you can see straight away with a level of confidence exactly how Watson has done that off the top of my head for anybody that wants to type this in now I think hes Gwen gwe and – system you dot my bluemix net were going play with that right now and you can understand.

the mood if your audience even if youve only got a small amount of people play with it on your own Twitter play with it on your company or your clients Twitter and youll have some fun with that thanks Shawn again thanks Jeremy its so second question and I think this was asked on the slide with the Gartner hype cycle so the question is you see a 28.

percent will integrate ai in their strategy for the next three years in which business I assume that its meant to be business is is it going to be all customer products metal industry machine manufacturer etc question mark yeah isnt that a great question because there’s nothing worse than somebodys standing at the front of the stage or in a webinar throwing out a generic number.

as if its going to apply to everybody I love that Winston Churchill quote the only stats you can trust are the ones you falsified just stealth I really like that we could make numbers mean anything marketers are great at that know the context for that was I went out to a thousand marketers across Western Europe in the UK these are senior marketers so really anyone that’s a head of all above.

director level VP CMO and they were from a whole cross-section there was a huge amount of financial services but it was a cross-section of b2b and b2c the fastest use cases are probably going to come from consumer products certainly CPG and one of the reasons for that is because you dont always have a direct relationship with the customer you think about a Unilever Procter & Gamble.

General Mills Nestle dont always have a direct relationship so you need to rely on anonymous data to try and understand what to do next with audiences whether that’s creating an advert you know whether that’s trying to understand the impact that your cons may have had worldwide in different markets really interesting for CP financial services by far theyre going to be the ones that jump on AI the fastest mostly because of the operational cost savings customer service for sure from chatbots Ive seen.

at the moment companies like RBS and Vodafone I cant give the numbers but dramatic reductions in cost savings much faster call times huge drops in the cost to serve customers and as a result increased customer satisfaction because now youve got a Bart they can understand natural language processing you know underline waiting for someone for a basic task youve got whats than understanding and having a real conversation and you’re able to answer.

those real-time questions super fast with a very very high level of accuracy that’s the reason I point out Project debater by the way because when you go and look at that on YouTube youll see how accurate and how compelling Watson is in understanding the way that humans respond customer service is probably going to be the biggest use case I think but as far as that data it really is a cross-section yeah 28% Im going to be adopting it over about 18 months some industries are going to be slower than.

others you know retailers like Nike up in jumping on it for quite a while coca-cola for sure but yeah if your work in financial services big opportunity telcos as well our customer service right thanks Jeremy so a slightly longer question here but but its a great question a little less could Watson social sentiment analyzer help us to find people ready to listen about how we might be able to help them increase.

their revenue and just the second part of that question Jeremy is either by crafting content on platforms like Facebook or in some way to actually find people like new customers mm well the easy answer to that is yes that’s when you’re looking at social sentiment and is why when you look at Watson whats in it really that is worth clearing got what Watson is because.

again he covers a multitude of sense what some basically is is a massive tool box of apis you know these application programming interfaces this its just a toolbox of stuff that helps you do things really fast people mistake it for this great big AI supercomputer that just you feed things in a magical stuff happens it takes a lot of people to make whats and work and it takes a lot of training with real data to make whats.

and work as well and something like social media sentiment is no different personality insights is a good example because you can feed a Twitter account in and it shows you straight away the sentiment and it starts to show you weve got sentiment analyze the tone analyze the personality insights for you different apis that show you the mood of the audience now where this is interesting is when you look at the difference between personalization and personification fact there’s a TED talk you could watch by one of our senior.

researchers is called Dario Gill dar IO Gill gee i double-l and what he does in his TED talk is he shows a real-life example of using insights based upon data to train fire and companies who you might want to buy basically hes looking at an audience that you might want to play with or deal with well think about the Alex de Kidd example all that likes the kid is doing is looking at millions and millions of people to try and find an unsigned.

artist that has exactly the same attributes tone of voice sentiment personality fit same language you know and he fed in a lot of language and a husk where words that train train Watson defined Rhapsody that’s all you’re trying to do across social now the issue is and this is why I paused the LA bet when you scraping social for some of this sauce stuff sometimes you’re looking at a very very small cross-section of an audience in fact Facebook is at the most its going to be 15% of an audience if you’re.

looking at Twitter each probably going to be less than 8% maybe even closer to 5 the majority of those customer conversations are taking place some private messaging apps that’s why looking at visual and voice 10 colleges so you apply that to something like Instagram for example like to read every single photograph on someones Instagram and understand the mood that they may be in and then back to Darios point Dario girl you could start looking the mood of an audience in real time so.

a bank could be thinking well an investor for example heres the product that we might want to sell this customer based upon what lives in their database that could have been taken from interest graphs and economic graphs so I could be taken from conversations youve had with them but the mood that they seem to be exhibiting at the moment through downloaded this one might pay put search for this thing posted this and Instagram said this on Twitter the mood that they are now in and where wasnt able to surface that based upon that Free.

Anonymous put with the available information it may be that theyre looking for a completely different product now the use cases for that could be anything that could be recommending your product it could be predicting based upon the mood of that person how many seconds before they leave the website right think about that someone lands on your website and weve got a pretty strong understanding that they may want to leave your website in 75 seconds so you might want to serve them.

a piece of content or personalized discount in 60 seconds to keep them on will start increasing conversion rates you can start surfacing the same language and customer service to make sure youve got a better satisfaction quicker there the literally is a thousands use cases that’s why its a very long discussion to have to Train answer a quick question but go on play on those apis and start looking at some of those things that theyre able to do those your sentiment is absolutely.

killer just remember that some of those social listening tools that weve been using for the last five years just dont work in the way that theyre used to because those conversations just dont really exist publicly and the way that they used to got Brinker calls it dark social if you dont know about it go and search that term you could say dark social the report that I published last year that goes into a lot more information about this and Scott brink are also wrote a book called hacking.

marketing which speaks to this in a lot more detail as well shows how things like Watson an AI can help exactly that self hope that’s useful check out those resources good question thanks very much so I think well probably leave this as the final question I realize we only got about two minutes left and we want to be respectful of course peoples time weve already taken an hour out of your day already so the final question and a great one actually nice and a nice and.

easy do you have a case of competitive analysis using Watson Im competitive and Im not sure what that means competitive analysis doing what do we know what that is so my understanding was that it was competitive analysis maybe across like different companies or different industries maybe maybe yeah let me let me Ill point you in two different directions then Ill point you.

back to the Dario Gil TED talk because what he does in the first 75 seconds of that TED talk is he shows something that we never usually allowed to show and that show is basically how Watson is doing competitive analysis but I think its about five different companies going to understand the differences between them to understand which ones to work with so that’s a great that in that use case is trying to figure out which company to buy by what you’re basically doing competitive analysis to try and find the differences and nuances between.

them another way that weve done competitive analysis is using something called the News Explorer API and what News Explorer does and then this is Ill be the link right after this if you follow me on twitter at Jeremy wait Ill tweet the link so you can go and check it out news explorer allows you to put in any topics any keywords any companies any words and you compare multiple words one company with another company and it shows you all the news articles that are currently taking place across the world and it shows how theyre linked together.

and you’re able to see patterns and correlations in the middle of that which also helps for example for with things like competitive analysis I sometimes use it for that there’s a lot of different uses of competitive analysis so those are the two I will leave you with anymore that I havent answered feel free to ping me on Twitter Jeremie way and I can answer them and point you the directions of the links and I think with that one are we going to share the slides with everybody yeah.

we will indeed may so anybody on the call if you just use the link you used to access this webinar if you come back youll be able to download the the entire saved video of today so everything will be there and whats in there recommended stuff just before you go is if your reading list isnt big enough Live 3 0 by max tegmark I think its only about 2 pounds 50 these two euros 50 as well on iBooks on Apple you.

can get it from Amazon fantastic book its available in paperback talks about the future of AI and the implications for humans and its a really easy read its well worth checking up so with that thanks again everybody like Sean said we love you we really appreciate your time they can hour out of your day in the hope it was useful and you got something.

out of it and we will see you next time thanks again thanks for everything bye-bye.


Comments

Leave a Reply

Your email address will not be published. Required fields are marked *