Last year at this time, an impressive group of data and analytics industry specialists answered a single question in preparation for the year to come:
“What gets you out of bed in the morning when you think forward to data and analytics in 2016?”
Their answers did not disappoint, so here’s an update to what industry advocates, colleagues, friends, and even my close competitors in digital analytics are thinking about as we move into the new year.
Jim Sterne
Founder, eMetrics Summit & Digital Analytics Association
There are two exciting things on the horizon in 2017, one technical and the other human. On the software side of things, we see the very beginnings of machine learning in marketing. More companies are giving it a try. By the end of the year, we’ll have some solid use cases and best practices. We’ll also have some funny, some embarrassing and some brand-tarnishing examples of how not to leverage this new capability.
On the human side, I see the marketing department as a whole relying on analytics more as a matter of course, rather than a curiosity. I see product marketing managers, campaign managers and merchandisers tackling analytics concepts on a deeper level in order to leverage the digital analytics talent already in place.
As more education is offered to the non-analyst about analytics, the relationship between those who crunch the numbers and those who make business decisions will improve and then flourish.
Elissa Fink
Chief Marketing Officer (CMO) at Tableau Software
What happens when more and more marketers get data-smart? Entire teams start to embrace a culture of analytics. And that’s what we’ll see in 2017.
More organizations will adopt self-service analytics at scale. Modern business intelligence — this idea of moving the power to perform analytics from the hands of the few to many — will become the norm. And data skills, once a nice-to-have, will become table stakes.
With the entire team collaborating with data, marketing teams will work smarter and innovate faster. They’ll see great results and help drive the business forward.
Stéphane Hamel
Digital Analytics Thought Leader and Faculty Chair, Digital Analytics, SimpliLearn/MarketMotive
Last year, you’ll remember I mentioned the ad blockers’ growth. This year, I would like to raise attention toward the new EU data regulation which will inevitably disrupt the martech/ad-tech space by preventing companies from using an individual’s data unless they have direct consent from the consumer.
As Emma Hall said in an Ad Age article, “Everything that invisibly follows a user (anonymously or not) across the internet will, from May 2018, have to pop up and make itself known in order to seek express permission from individuals.” This is major, and non-EU vendors will be impacted, too. Companies building a good, mutually beneficial customer relationship and using first-party data will have it easier, but those heavily relying on third-party data will be challenged.
Back in 2009, I created the Digital Analytics Maturity Model — ages ago in Big Data time. I have developed a keen interest for understanding the transformational aspects brought through digital analytics. A couple of months ago, I started asking colleagues, other industry veterans and conference attendees, “Tell me, just between us, how is it really going for you?” As I suspected, the answers baffled me.
As an industry, we’ve been doing web analytics (now digital analytics) for a good 10, 15 years. We’ve been told executive sponsorship was essential to success, yet very few analysts I talk to actually have this luxury. Marketers and analysts alike thought they had a free ticket to the management table. This enticed me to write the Radical Analytics Manifesto. If we’ve been doing it in a certain way for so long without much value, maybe it’s about time to try something else?
There’s a link between those two stories: a disruptive tidal wave which will lead to a repositioning and questioning of the true benefits and expectations stemming from data, Big or Small. 2017 will be the year we go through the “trough of disillusionment.”
There are signals pointing to the realization many things are wrong with digital marketing, digital analytics and data science in general. Be it how most industries are “nowhere close to realizing” the potential of analytics, the “data science delusion,” how “attribution is impossible” and many others, in all cases, the root cause isn’t technology nor is it data; it is “people”: bridging the gap between various teams; difficulties in embracing modern, agile methodologies; and lack of discipline in “eating their own dog food” by measuring the value they bring to the organization.
Alex Langshur
Founder & Senior Partner, Cardinal Path
Last year I said the that the word was “activation.” I’m going to double down and say it again this year. But here’s why:
Awareness of the value of data management and data governance as a core 21st-century corporate need. The biggest issue with activation is how hard it is to collect, connect and manage data. Whereas in the past I found myself spending a lot of my time educating on this point, now I’m spending less time educating and more time responding to requests for support. I like to say that this is a 21st century enterprise problem, and organizations are waking up to its strategic importance and value.
Maturing of the data management platform (DMP) vendor market. The market has reached a tipping point in the last year as awareness has grown and consolidation has followed. Prices are in flux, functionality and ease of use is increasing, and importantly, there’s growth in the ecosystem of third-party integrators. We expect further changes in the market, including more consolidation, along with new startups.
The rise in programmatic. Despite some recent hiccups (including those that were brought under bright focus as a result of the recent US election), I believe that the growth in programmatic spend will continue. Better targeting of this spend will demand improved data management, which in turn means more and more finely tuned activation.
The holy grail of the consistent cross-channel customer experience. At the far end of the maturity spectrum are those organizations that have an audience blueprint — in other words, they are using their DMP to activate and target across the entire customer journey in a consistent and nuanced fashion. And the key word here is nuanced.
Gary Angel
CEO at Digital Mortar
I did pretty well with my predictions for 2016 — especially that whole “machine learning” thing I thought would take off. I feel like I should be on the radio, touting my Monday Night Football record and pimping my inside information! Sadly, those guys are proof that past history is no guarantee of future performance. Make enough predictions and some are bound to be right. So here’s more…
I’ll focus mostly on retail since my new business is all about measuring customer journey in the store. Speaking of which, building a business focused on a new market is the ultimate “I actually believe it” prediction. And what I believe is that digital methods are coming to physical retail and that omni-channel experience is the only defense traditional retail has against Amazon. If you’re trying to do real omni-channel, it doesn’t make sense to have all this detailed data about what people did online, what they’re interested in, what they thought about — and then match that up with nothing more than what they bought in store.
There’s more to the in-store experience than purchase and return — and I think 2017 is the year that omni-channel goes beyond transaction and becomes experience-driven.
And speaking of those in-store experiences, I think more than a few are going to be highly customized digital journeys that take actual advantage of being in the store. VR, social, and interactive digital experiences win at least a few in-store battles in 2017 and become the stuff that the next generation of retailers build around. That’s a fascinating area for digital analytics folks to engage with and drive — a combination of gamification and e-commerce should make for pretty interesting new measurement opportunities!
Daniel Waisberg
Analytics Advocate, Google
When I was asked what got me excited about data and analytics for 2016, I wrote that it was the fact that departments would increasingly talk the same language through centralized data analysis platforms. I believe that even companies that are still not doing it know that they should, so let’s call it old news!
Now that data has its place and a common language, comes the challenge of letting it be heard. With the recent launch of Google Data Studio (and an expanded interest in dataviz), I believe that in 2017 professionals will increasingly use visualization techniques to understand and communicate data more effectively. I am excited to see our industry moving towards more beautiful and actionable data stories!
Joe Nunziante
Group Director, Client Services at Cardinal Path
I think we’re due for a revolutionary new approach to digital ads — at least I hope we are. Many of the changes that we’ve seen over the past few years (native ads, cross-screen ads, ads powered by DMPs) are really just new takes on an outdated ad model. Display ads don’t work? Alright, let’s move them around and make them appear like content… I don’t know how much longer that sort of an incremental approach will suffice.
A key challenge here is that the entire ad buying ecosystem is built around a set of transactional terms and measurement approaches that keep us tethered to the past. For all the flaws of a CPM-based ad buying model, it’s fairly simple and makes for clear comparisons between media options. A new ad model may require a whole new set of accompanying metrics — and that’s a battle our industry will have to be ready to face.
Robbin Steif
CEO, LunaMetrics
As the “digital years” progress, we face an interesting phenomenon: Early adopters and forward-looking companies are experimenting with cutting-edge technology, while many companies are still trying to get clean data. The spectrum of what companies need is wider, and so for 2017 (and beyond), we consultants need to have a bigger bucket of tools. We also need to be more flexible, flipping from Big Query to straight engineering and back again.
Looking ahead to 2017, we’ve seen clear patterns emerging, not just in the work we do with existing clients, but also in the types of requests that we are receiving. Later adopters are starting to recognize the value of these data-based tools and internally are putting dollars and resources towards getting them set up. That kind of organization will expect more out of their data in 2017 — the kinds of things we were talking about in 2006, i.e., more insights, more actionable pieces of information.
Moving along that spectrum toward the still-not-new but slightly more interesting end: The process of integrating tools has started to become easier, but this will continue to be a focus in 2017. Every tool has their own analytics and reporting; won’t it be great when these are consolidated in a single system and similar methodologies? Developing a source of truth is crucial, whether that’s internal data warehouses or a common system like Google Analytics.
Lastly, I think storytelling will be more important than ever. What does our data tell us about our customers and our visitors, and how can we use that information? Analytics is a part of every department, and when companies start recognizing that, they’ll be more likely to reap the benefits of a solid implementation and greater understanding.
Juan Damia
CEO & Founder Intellignos — Endeavor Entrepreneur
As we wrap up 2016, a brand-new year is waiting for us, and we’re still thinking about what we’re going to do to solve all our data problems. Organizations are still suffering from unreliable data, bad platform implementations, overwhelming reporting, lack of data integration, and difficulty converting data into decisions that impact the bottom line.
In Latin America, less than 15 percent of companies are able to turn decisions into action. The question is, if we all agree that making decisions from available information is not getting as simple as it was supposed to be, how can we solve that situation in 2017?
1. Governance: Analytics governance is still as important as last year, even though we haven’t seen the huge effort from companies towards the development of an analytics governance department. Hiring a data scientist won’t make your company change its culture. It has to be a clear requirement from the companies’ management flowing down through the entire organization, and it’s important to consider external experts that can prevent a company from getting stuck in its current culture.
2. Automation: Machine learning has become a great solution enabling the automation of those 80 percent of activities that generate 20 percent of the revenues, allowing managers to focus on the 20 percent of activities that generate 80 percent of the revenue. This is where the value lies, and it requires (and justifies) that in-depth data digging and analysis. Creating and optimizing clusters that can be pushed into a DSP [demand-side platform], generating predictive models for marketing automation, or understanding the best action that can convert a micro experience into revenue are some of the urgent things that machine learning and big data analytics can simplify and allow us to focus our efforts on the more important things.
3. Dashboards vs. Insight Outputs: The volume of data [and] shortened decision-making cycles make it really hard to have enough time for interpreting information and making decisions, and analysis and reporting are two completely different things. Presenting a stakeholder with charts and graphs leaves room for interpreting the data incorrectly or without the proper context. Machine learning techniques and data storytelling allow specific insights, explained with the relevant context, providing real value to stakeholders.
Simo Ahava
Senior Data Advocate at Reaktor
Last year I was looking forward to a paradigm shift. With all the major advances in technology, particularly in scalability of data storage and processing, it seemed odd to me that we’re still so fixated on single-vertical analytics (web analytics, app analytics, business analytics) instead of something more comprehensive.
And then it struck me. It’s not that the industry isn’t moving forward; it is. The problem is that we still struggle to conceptualize complexity in such a manner that would make sense from a knowledge transfer point of view. If we can’t consistently introduce change into the companies, organizations and societies we work in, how could we expect to do it with our data processes, too?
So, my prediction — nay, my wish — for 2017 is that people working in digital take a good look at their current skill sets and re-evalaute them from the perspective of, “What do I need to learn in order to drive my business forward, especially in terms of using data to fuel the change?”
For many, this would mean that they’d need to dust off that “JavaScript For Dummies” book, or finally take that Python course at www.codeacademy.com. For some, it might mean revisiting the basics of statistics. And for everyone, a good look at the business logic of a data-driven process is warranted.
Data and analytics don’t exist in a vacuum; they derive meaning from their application and their context. A set of good data can be destructive when in the hands of an unskilled practitioner without the passion to understand what the data is for. And that same data set can be constructive when in the hands of a hybrid learner who has taken the time to amass the skills required for working in analytics.
Finland, the country I am from, is turning 100 years old in 2017. It’s a chance for all of us Finns to re-evaluate what it means to be Finnish, and how we negotiate the past, present and future, to make this country something our children would be happy and proud to live in. In 2017, data will be as old as the known universe, so it’s high time we do the same with data. We need to catch up with the speed of change and re-evaluate our own skills, especially in the context of the businesses, organizations and societies in which we live and work.
Corey Koberg
Founder & Senior Partner, Cardinal Path
As marketing analytics becomes more customer-centric, the need for canonical customer profiles becomes critical. Will the CRM step up and take its rightful ownership as the authoritative customer record (and “attach” attributes from other systems) or will the martech/ad-tech industry continue to innovate around them and supplant?
Marketing analytics long ago moved from the CIO/IT world to the CMO’s purview. With privacy and regulation becoming too costly to ignore, will the Chief Legal Officer become the new analytics owner in big business?
Simplistic demographic based look-alike modeling (males, aged 25–34, etc.) has existed long past its expiration date not because it works, but because it’s easy to buy and easy to scale. There are better ways to crack this nut, and brands are demanding they be put to use.
Behavioral modeling needs better visibility into causality. Just because ice cream sales and drownings go up in the summer doesn’t mean I can prevent drowning by banning ice cream, and similarly, just because two behaviors are correlated doesn’t necessarily mean I can drive more sales by encouraging those correlated behaviors (or even finding more of them). We need to employ the data science that helps us understand the *why*.
TV isn’t going away and online/addressable is growing, especially with cord-cutting trends. “TV” (broadcast video) won’t go away, but the merger is coming and with it a whole new set of metrics. TV buyers love standardized measures above accuracy (Nielsen ratings, etc.). I’ll be curious to see if that mentality carries over in the brave new world.
Feras Alhlou
Author, Co-founder & Principal Consultant E-Nor
If we said that multi-channel and multi-device attribution will be a tough challenge for marketers and analysts, that would be old news for 2017 (but still very relevant). As an industry, we’re certainly still working to derive insights from user experiences across multiple channels and devices, but we’re now also increasingly confronting the imperative of analysis across multiple datasets. Marketers, analysts and business owners must brace for the plethora of data sources that will be readily available and at relatively low cost.
The Extract Transform Load (ETL) model is as relevant as ever: Systems still don’t talk nicely to each other, and it is still up to you to transport, clean, stitch and aggregate your data. And we might as well add the V in ETLV (for Visualize). Once you’ve performed the first three steps of the process, you’ll then be able to chart and navigate your data to mine for gems that reveal the subtle insight (or glaring anomaly) leading to significant improvement (and a promotion for you in the new year).
All the above will require commitments to systems, processes, people, mastery of core implementation and analysis skills from the start, and of course budgets, but if you can begin to score wins across the multiplicity of devices, channels and datasets, you will have laid a foundation for a sophisticated analytics practice and competitive advantage through data.
Matthew Tod
Director, D4t4 Solutions
So what excites me this year? Well, I have had my first forays into using machine learning, artificial intelligence and neural networks in the last 12 months, and the initial results excite me greatly. Also, it has revealed that I need to focus more on the collection of experiential data, not just behavioral data.
Experiential data is all the information present during an interaction but not actually clicked on. For example, it includes the availability of a product in my size and color, or the choice of flight times offered to me, or the real savings I am experiencing in £/$. All of this rich data is great food for deeper analytics and smarter personalization, provided of course it is collected.
So in 2017 I’m going back to basics — collecting the experiential data I really need to make a commercial impact.
Allan Wille
President & CEO, Klipfolio Inc.
Millennials are starting businesses at twice the rate of boomers. They are also rising quickly to positions of corporate leadership. These socially minded, digitally native, entrepreneurs have different needs, expectations and requirements for data and analytics solutions that are going to drive three trends in 2017:
Increased demand for cloud integrations: Since Millennials were born in and live in the cloud, they think cloud-first for their business applications. We have customers who no longer have any on-premise data — not even spreadsheets. The increased demand they are driving for cloud applications will, in turn, accelerate an increased demand for analytics tools to have more, and better, cloud integrations in 2017.
Emergence of wallboard transparency: Millennial entrepreneurs and leaders are taking the same transparent, share-all view they have in their personal social media soaked lives, into their work life. The trend that is emerging from that is what I call “wallboard transparency” — the display of performance metrics and data (even financial) on wall-mounted TVs for the entire company to see. I expect this trend to accelerate in 2017 as more and more Millennials rise to positions of leadership.
Increased demand for real-time metrics: Millennial entrepreneurs and leaders were weaned on Google and the instant answers it provides. When they are presented with the current way of looking at performance data once a month or quarter, it seems painfully outdated to them. They want to continuously know how they are performing. This will continue to drive the demand for real-time access to business metrics.
2017 and beyond
As the digital data and analytics space continues to evolve, old problems will be solved and new challenges will be uncovered, providing us all with the next great opportunity to leverage the power of data. Whatever your digital goals are for 2017, I think we’ll all agree that we hope that data and analytics will play a part!
Comments