“Data sucks.”
Her two words started our meeting and ended our pitch. The ad agency exec didn’t stop there.
Her quantum prick at the insight balloon became an unstructured hole you could drive your biggest data through.
“The whole thing is a mess. Data should drive our plans. Really, it can’t. Vendors come in here and tell me how many numbers they’re going to give us. In the end, our analysts just get a bunch of who-clicked-where tables. Fast forward a few months and we get charts and even more tables. Most people can’t or don’t have the time to act on these so-called ‘insights.’”
Then a week after our meeting, WPP sold off most of its data business Kantar to “focus on data and digital.” Clearly, this is more than rebranding; the data world is changing.
What’s an Insight and Where Do I Get One?
Turns out the what is the easy part. An insight drives your brand. Apparently, creative types know this and keep it to themselves.
“Why is a good insight like a refrigerator?
Because the moment you look into it, a light comes on.”
– Jeremy Bullmore, WPP Advisory Board
No, you don’t need an insights fridge. A friend and expert tells us that all we have to do is, “Ask.”
Our expert is Ashish Agrawal is an experienced commercial analytics lead. He builds teams that build data warehouses that build reports. His advice, “Stop thinking about reports. At the end of the day, fundamentally, people ask me a question.”
Like our agency exec, his brands have lots of numbers and lots of pretty charts. Those things tell his marketers how they did. Emphasis on the past.
His marketers ask questions to guide what they will do. Emphasis on the what’s going to happen. Answers—not charts—drive strategy.
In Ashish’s Utopia, every Monday, his people would ask, “What do I need to know—right now?” A big part of his job is to get people to ask the right question.
“Our industry loves complexity. Veeva changed/simplified CRM. Right now, time lag is a real problem. Data is a real problem. Presenting data is a real problem. Our mindset has to change. It’s time to change data.”
An Insights Arms Race
IBM spent more than $15 billion to build Watson. In June, Google and Salesforce spent $18 billion to buy Looker and Tableau, respectively. Here in healthcare, the private equity fund, Bain, bought most of Kantar.
The investment bank, Oaklins DeSilva+Phillips, calls it an “insights revolution.” They ran an event to discuss it in early October. Managing Director, Ken Sonenclar, summed it up like this:
“Clearly, the enormous investment to transform old-line market research companies into data and analytics powerhouses signals the huge opportunity for those who can generate real insights for their customers.”
The recurring themes at Ken’s event were that insights are:
- Easy to consume.
- Smart.
- Self-serve.
- Unique and lead to real-time decisions.
Leaders from Kantar and Bain spoke at Ken’s event. Their history, scale, and being at the top of a mountain gives them unique perspective.
“Market research is the means. Insights are the end. When we started, there was no data out there. Now, there’s too much data out there. Our clients need to parse through it to find the value.”
Value = Data + Context
Context is the key. IBM showed us how this works in real-time at the U.S. Open tennis championship.
On our behalf, IBM asked:
What highlight should I watch?
IBM’s AI platform, Watson, ranked each point using score, crowd noise, players, and court. The points that mattered most were the highlights you should watch. Like AI for color commentary.
Here’s how context plays a role. Let’s say the score is 40-love in the first game of a match between two unranked players. Probably not a point anyone cares about. Now, add that the crowd went crazy. That context made the point a highlight worthy of ESPN’s SportsCenter.
Or match point on center court between Bianca Andreescu and Serena Williams. The import of the situation makes the point a big deal.
Like IBM’s Watson, John Almeida is big on active crowds. At Ken’s event, John talked about “engaged data.” Instead of using lots of data with lots of zeroes, he says use smaller datasets full of results.
John recently invested in the pharma survey company, InCrowd. InCrowd has delivered 2.5 million answers for more than 300 pharma brands.
Don’t kid yourself. Datasets are not perfect or complete. For years, we’ve used sparse data to model reality. Heck, the clinical trials that underpin our entire industry are samples.
The art of insight science is getting a good proxy dataset—quickly.
Answers and Insights – At Speed
As we race to personal marketing—at scale—we need answers—at speed.
A funny thing happened at Ken’s event.
In the breaks, everyone pulled out a smartphone.
They fired up Facebook.
They checked Gmail on their iPhones.
They saw if Amazon delivered their packages.
They texted friends.
They called their offices.
The insights we want are right there in plain sight.
Which companies know us best?
Google, Facebook, Visa/Amex, Apple, and Amazon. That three of them dominate the ad world is not random. In real time, they know what we read, view, and buy. Plus, who we text, email, and care about.
The things they know are the insights our brands want. But they only let us access them by buying ads on their platforms.
Kantar told us pharma is flush with data. They’re right. More than one million stories hit PubMed each year. Tennis points. About 200,000 doctors use Twitter to tell you what they want. Crowd noise.
What if pharma had a machine that could ingest and tame that largess?
Brands would have a fast, easy, self-serve, color commentary.
Good news.
This answer machine isn’t theoretical.
It can already answer pharma’s traditional and new questions.
So, What’s Pharma Asking?
We asked some pharma folks the questions they most want answered—and then looked at how an insight machine can now help them to answer them.
Anna – VP Media Director
Q: How should I allocate budget across publishers for a very narrow indication?
Her ask is more nuanced than, “What’s the best way to reach a diabetes audience?”
The old way was to match lists with a vendor and ask how many were regularly active. With an insights machine, she peers right into a small community. She can ask who the KOLs are and which stories drive the market. Now, she can see which publishers best meet her needs.
Jessica – Multichannel Integration
Q: Am I winning the battle for clinical trial mindshare?
Her brand is in a two-horse race to win a market. At a big meeting, she’s releasing results from her clinical trial. Did doctors move to her brand?
Without an insight machine, she sent people to the conference. It cost her group money and lost time. She could have paid an outside group to do it. That would have cost her even more and the report would have taken even longer. At her big meeting, an insight machine gave her actionable insight daily.
An insights machine can answer questions we couldn’t ask before. Or ones we couldn’t ask easily. That lets brands adapt to changes more quickly or create new strategies.
Xavier – Associate Director, Promotion
Q: When is my audience really tuned in?
This is a new question for pharma. “Tuned in” has been a big thing in TV for decades.
Did you watch the Friends finale? Yeah, everyone else did too. That night almost everyone watched TV. And nearly everyone who watched TV watched “The Last One.”
Ratings is a slice of the pie chart. Tuned in tells you how big the pie is. Xavier might know his slice. Knowing how big the pie is tells Xavier when to put more media into market.
Doug – Global Brand Director
Q: Are my KOLs with me or pulling a Kendall Jenner-Pepsi thing?
Doug invests in KOLs to build his brand. Then he saw Kendall Jenner promoting Coke after Pepsi paid her a bunch of money. He began to wonder, are our influencers committed to our brand?
It would take a big analytics team to answer this. Most brands don’t have that kind of horsepower. An insights machine levels the playfield. It can check social media.
Read bylines on peer-reviewed papers. See which KOLs were presenting any competitive papers at conferences.
Bottom line: Data sucks. Insights don’t.