23 min read July 11, 2017 Andrew Park Is Machine Learning the Future of Marketing? Experts Weigh in.Why do 97% of marketing influencers believe the future of digital marketing will involve human marketers working with machine learning-powered automation? Thought leaders in PPC, social and mobile marketing explain in this in-depth survey.Where is digital headed?Are disciplines such as search engine marketing, social and mobile marketing all trending towards a fully automated world where artificial intelligence (AI) robots take over our jobs? Not necessarily. In a survey of top influencers in online marketing with expertise in paid search, social and mobile, 97% of respondents suggested that the future of marketing will actually be smart marketers working hand-in-hand with machine learning-based automation solutions.To help us understand the growing role of machine learning in marketing, we spoke with some of the top influencers in the space, including Michael Brenner (@brennermichael) of Marketing Insider Group, Serena Ehrlich (@serena) of BusinessWire, Adelyn Zhou (@adelynzhou) of TOPBOTS and Chris Messina (@chrismessina), creator of the hashtag – among others. Selected responses to the survey are below.Interested in learning more about how new technology and academic disciplines such as machine learning and data science are changing digital forever? See our full report on the intersection of data science, machine learning and digital. Joe Martin @Joedmarti – Social Analytics Lead & Evangelist, Adobe“I see the future of marketing as a combination of smart people & smart automation. Robots led by people. I think it will put a greater emphasis on data storytellers vs. data scientists. Of all the marketing technologies using machine learning and AI today, programmatic ad buying and improvement of ad spend allocation through machine learning impress me the most.” Serena Ehrlich @serena – Director of Social and Evolving Media, BusinessWire “The future of marketing will be some combination of smart people and smart automation technology. Why? That’s the next step in the process. Data + experience.” Travis Wright @teedubya – CTO & Co-Founder, CCP Digital | Author, Digital Sense“Marketing in the future will be a combination of smart people and smart automation technology because machine learning and AI are able to help with lots of our menial, mundane, and repetitive marketing tasks. This will, in theory, free up marketers’ time to be more creative and focus on big tasks.Every week, on the VB Engage podcast, we encounter new startups that use AI & machine learning to automate various tasks. The future is here, folks.” Phil Gerbyshak @PhilGerb – Chief Digital Officer, Vengreso“How do I see machine learning affecting search marketing, paid advertising and intent-driven channels? It will lead to mass personalization and super-specific paid advertising that knows our intentions even better than we do.” Stephane Kasriel @skasriel – CEO, Upwork“We use natural language processing to create new SEM campaigns based on keywords that come from the user-generated content on our site. We use machine learning algorithms to decide on promotions we should be offering, on emails we should be sending and on [the] type of retargeting we should be doing. I think that AI will be able to tweak creatives and/or pick the winning creative, and similarly, it will be able to improve targeting (e.g. keywords on SEM, email addresses in re-marketing campaigns) but will not be able to create target segments by itself. “ Rick Mathieson @RickMathieson – Creative Director, iMATHIESON.com“[I work] with clients that develop machine learning- and AI-based solutions for brands for all sorts of applications, from personalization to conversational [chatting], to customer service and more. The low-hanging fruit comes in automated data visualization and the ways we can absorb and assess real-time data. What’s really amazing to me is some of the technology for context analysis — though as I said, it depends on a large enough data set and inputs. It also needs to assess ever-changing associations between consumers and their devices, networks, locations, personal or professional lives, purchase histories, historical and real-time behavioral patterns…and more. Some of the technologies are getting pretty amazing — approaching that Promised Land, or Holy Grail of marketing. It’ll be exciting to see that happen.” Pek Pongpaet @pekpongpaet – Managing Partner, Impekable“I see the future as some combination of smart people and smart automation technology, because AI will need learnings from humans, or perhaps fine-tuning. Marketing is messaging and is both an art and science. I believe some aspects of marketing are all automated as they already are, but as it stands, it does need humans to set that up. “ Michael Brenner @BrennerMichael – CEO, Marketing Insider Group“Of today’s marketing technologies using machine learning and AI, natural language processing (NLP), chatbots and voice-driven search impress me the most.” Albert Renshaw @Valuable – CEO, Apps4Life / Minty Networks“I think with deep learning and proper tracking, searches will become entirely customized to the individual user, further than just what results they see, but even the whole experience itself. That is, page layouts, graphics, UX etc. Some people respond better to banner ads, some people respond better to pop-over video ads, some people respond better to X. Everything will get personalized and optimized for your unique personality type. Google AdMob is experimenting with some cool advertising technologies, which, from what I understand, track user behavior after they install an app from an ad. They work with the app developers to help identify key actions in the app that are considered ‘valuable,’ and then find out which types of humans are most likely to complete the highest number of these actions. We’ve moved from cost-per-impression and cost-per-installation to cost-per-post-installation-actions. We are currently in contact with a team at Google about getting these types of ads integrated into our apps.” Joe Chernov @jchernov – Vice President of Marketing, InsightSquared“Machine learning is best for exposing what people don’t know that they don’t know. It also benefits greatly from larger data sets. Those two strengths converge in areas of search marketing, paid advertising and internet-driven channels. I see machine learning helping the manager of these channels make more-informed decisions, the way traffic-sensing map apps help a driver take a smarter route. AI still appears to have limited penetration in B2B marketing.” Adelyn Zhou @adelynzhou – Emerging Tech Strategist | CMO, TOPBOTS“We’re working with brands who are using these advanced techniques in a few ways: 1.) Customer experience – This is in the intersection of marketing and customer service. Brands are using AI to route messages better, respond to questions faster and provide improved responses to customers.2.) Visual search tagging – E-commerce companies are using AI-powered systems to better categorize and tag products in their catalog. If a customer searches for ‘floral sundress,’ the system would be better at finding all relevant results. 3.) CRM database management – To send more-targeted and personalized messages to existing users in their database. For example, based on a user’s drink preference matched with the weather in their area, a system will send a coupon for a Long Island Iced Tea. While some of this could have been done before, with machine learning, it’s being done at a deeper and more-personalized way. I’m sure 90% of people said that the future of marketing is a combination of smart people and smart automation, right? While we will move towards automation, there will always be an element of human judgment to it. The human CMO, hopefully, will create the strategy with input from data, models and longer-term AI recommendations. Marketing is part art and part science. While both of those fields are being disrupted by AI, there will always be an element of human in judgment. We’re already seeing algorithms automate the majority of paid online ad buys. The programmatic purchasing of advertising is just one step away fully automated machine learning systems. I see more and more of these channels automated. And we’ve seen how automation allows marketing messages to be personalized at scale, quickly and at a lower cost. I’m excited about the visual recognition tools and how they are able to tag images and video. We can now understand how much brand awareness and “media value” is derived from user-generated content or on a TV showing of the NBA finals by processing how long and in what context a brand product was shown. Longer term, I’m excited for AI to help copywriters generate better taglines and email copy personalized to the user level.” Jed Schneiderman @jedschneiderman – Co-Founder/President, Tapped Mobile“The future will be a combination of smart people using smart automation technology. This is because people interpret results. People are not always predictable, and it’s hard to always incorporate market forces and other variables. Marketing will become more return on investment (ROI)-driven, less human-focused and more efficient.” Andy Zimmerman @ahzimmerman – CMO, Evergage“Serving up ads across digital channels that are truly personalized to the individual or account (based on attribute, affinity and intent data) can only be done using machine learning. And it’s well worth the effort to do this, because 99% of ads don’t generate a click because they’re irrelevant or ill-timed. If the data is available about your target audiences to help you deliver a message that will resonate and generate even a modest improvement to your CTR, it’s easy to make the business case.” Tom DeFloria VP Digital Solutions, InfoGroup Targeting Solutions“We use machine learning for our corporate training. Namely, each year, we have to pass a data security exam after taking an online course. In my opinion, the future of marketing will be a combination of smart people and smart automation. Why? Technology is providing us the ability to track more data in a real-time fashion. Analytics are enabling that data to transform into usable output information. Smart people are now able to spend their time/talents on reviewing the output and making informed inferences to complete the entire picture. Machine learning will continue to synthesize the vast amount of data/impressions online daily…billions of transactions formulate a picture of each consumer, buying patterns and web optimization.” Manish Patel @manishologies – CEO, Brandify“At Brandify, we use natural language processing (NLP) capabilities to increase digital engagement late in the funnel. When our clients’ customers want to understand the purchase they are considering, want to do further research and need help in finding the right information, voice search and NLP gives them answers that are more relevant to them in a quicker and more-efficient way. In addition to NLP, we provide Association Rules analysis – a rule-based method to discover interesting relationships among variables in our database – to provide these individuals with more-helpful and relevant recommendations.Overall, we use Machine Learning to spot patterns in ratings and reviews to speed up brand responses to customer feedback (across the positive, neutral and negative scale). Because reviews and online recommendations have such an impact on local engagement and conversions, machine learning is primarily used to quickly respond to local customer feedback and amplify the voices of our clients’ biggest brand advocates.” Erik Harbison @erikharbison – CMO, AWeber“Today, we’re using machine learning, deep learning and AI in our marketing by running tests using chatbots via Facebook messenger ads aimed at generating more-interactive discussions during the consideration phase of customers’ journey. I see the future of marketing as a combination of smart people and smart automation. This is because the percentage of people vs. machines will fluctuate towards, and most likely favor, machines in the next five years, but there will always be a need for human-only capabilities in the development of these machines.[Users will] eventually be getting hyper-targeted, showing you the right ads before you have the need. Something Amazon should be better at given the depth and breadth of data they’re collecting. Also, chatbots are appealing and have some room to improve on future iterations (such as in terms of UI and complexity of sequencing).” Sarah Wallace @sarah_wallace – Director Service Enablement / AI / Analytics, IHS“The future will involve some combination of smart people and smart automation technology. Why? For the near future, we will still need ‘machine teachers.’ These channels will still be using analytics, just more-sophisticated/self-learning data models that come with machine learning.” Roland Smart @rsmartly – VP Social/Community Marketing, Oracle“I am not actively using machine learning in marketing today — I am investigating the solution ecosystem. I’m particularly interested in AI to support social messaging services (e.g. Facebook Messenger). However, the future will be a combination of smart people & smart automation technology. It will absolutely be a combination of AI and humans working together. AI will allow us to dramatically improve human productivity by focusing effort on the highest value tasks. Machine learning will have a significant impact on intent-driven channels such as search and paid, not only on an optimization level, but also because it will create new contexts in which advertising will be deployed. However, to be frank, there are very few services that I’d position as including genuine AI today (though I am following a couple that are in stealth mode). There are more (like Oracle) that leverage machine learning to optimize programs. A lot of what I’m seeing today is focused on automating low-value tasks so that humans can be more productive. Then, there are more-advanced new applications (not existing apps enhanced with machine learning) that are really purposely built to work alongside humans…to give them marketing superpowers. Finally, there are a handful focused on fully AI-driven experiences that are purposefully designed NOT to be human….there are cases in which a user prefers to engage only with the bot. “ Stewart Rogers @TheRealSJR – Director, Marketing Technology – VentureBeat Insight“The future of marketing will be a combination of smart people and smart automation technology. This is because AI lacks emotional intelligence, and the buying process is still driven by emotion. Of all the machine learning solution out there, I’m most impressed by creative AI tools, various image recognition tools (and the smart use cases they’re involved in), Likelihood, Watson Ads and various SalesForce automation tools.” Shawn Goodin @shawngoodin – Marketing CTO, JPMorgan Chase“We’re using machine learning in our marketing to build propensity models to drive personalized customer experiences. But the future of marketing will be some combination of smart people and smart automation. This is because, as we move to cross-channel, cross-line-of-business (LOB) campaigns, automation is required to scale. We will always need to be interpreting campaign results to drive new insights to inform creative, branding and tweak experiences. As for how I see machine learning affecting intent-driven channels such as search and paid, data is the name of the game. As we get more cross-channel data, we will build more-personalized customer experiences that will continue to refine audiences…content by channel in context.” Bob Pearson @bobpearson1845 – Chief Innovation Officer, W2O Group“We use machine learning to build training sets related to our analytics work (influencers, content, language, channels). We are also exploring how to use machine learning to build training sets related to images to help the creative process. Both are in early stages.I see the future of marketing as some combination of smart people and smart automation technology. Why? AI/machine learning/deep learning scales human judgment. We (machines and humans) are better together. Both of us are sub-par alone. How do I see machine learning affecting intent-based channels such as search marketing and paid? It will completely change each model. We are already doing this in SEO, channel selection and more. As for which machine learning-based solutions today impress me the most? Great question, don’t have a great answer, which is due to the infancy of this area. Many folks say they can do things that they can’t. We rely on Google as an aid and other firms we find per project. Area cannot mature fast enough.” Summer Gould @sumgould – President, Eye/Comm Inc.“The future of marketing will be some combination of smart people and smart automation technology. Why? People should always have an eye out for issues that can come up. How do I see machine learning affecting search marketing, paid advertising and other intent-driven channels? I see them being an integral part in the future.” Bernie Borges @bernieborges – CMO, Vengreso“Machine learning will eventually do most of the tedious research. People will still need to provide context and monitor the relevance of data being examined.” Helen Keegan @technokitten – Founder, Heroes of the Mobile Fringe Festival“Search marketing is a shoo-in for machine learning, but people will still be required to account for any changes in the search engine rules and to counter the fact that everyone is doing the same thing.Insight gained from ad analytics and profiling should make paid advertising easier and more accurate. However, the analytics I’ve seen are patchy at best. The numbers don’t tell the whole story. Plus it means we miss out on the human element. As for which current solutions impress me the most, Facebook and Google currently reign supreme when it comes to machine learning, data crunching and generating income and clicks. I don’t think there is any company using true AI for marketing.” Tamara McCleary @TamaraMcCleary – CEO, Thulium.co“We are using AI in our organization on our website to learn what a visitor wants, and direct them to a more-personalized experience of our website. We are also currently exploring the use of an AI chat bot on our organization’s website as well. The future of marketing will be a combination of smart people and smart automation technology Why? The billion dollar question! The sweet spot is found when you combine the speed and intelligence of AI with the creativity and intelligence of human beings. How do I see machine learning affecting search, paid and other intent-driven channels? Better segmentation of audience, increased personalization and increased conversions through better lead generation. As for which marketing technologies using machine learning and AI today impress me the most, I like where Watson is headed. We use Sentient Ascend and Boomtrain, but there are so many exciting products coming to market right now that it’s like being a kid in a candy store and very easy to become distracted!” Ayelet Noff @AyeletNoff – Founder / CEO, Blonde 2.0“I see the future of marketing as some combination of smart people and smart automation. Why? Because marketing/PR will forever be a relationships game – not only automation. You need the right connections with writers in order to make your earned media a success.” Esteban Contreras @SocialNerdia – VP Product, HYP3R“How are we currently using machine learning, deep learning and AI in our marketing efforts today? Our software, HYP3R, leverages this to turn social media management into a recommendation engine. I see the future of marketing as some combination of smart people and smart automation technology. This is because it is simply not possible to do one without the other. As for how I see machine learning affecting intent-driven channels such as search and paid, the answer lies in results-oriented pattern recognition.” Chris Messina @chrismessina – Organizer, Botness | Advisor, Tagboard.com“I see the future of marketing as being a combination of smart people and smart automation technology. Why? AI is essentially all about probabilities. While you can do a lot with that, you still need human creativity to craft effective strategies and narratives as well as accounting for emergent cultural behaviors and societal context. As for how I see machine learning affecting search, paid and intent-driven channels – it’ll open up new targeting contexts (i.e. in the real world) and help improve the accuracy and effectiveness of targeting. And as for which technologies using machine learning and AI today impress me the most? Self-driving cars.” Thanks for reading our survey on top influencers’ thoughts about the future of machine learning in marketing. To learn more about how cutting-edge marketing technologies and strategies such as machine learning and data science are changing marketing today, please see our full report.