What I Learned from 5 Years at Gartner

Last month, I put this article up on LinkedIn Pulse and went on with my day. The response was extraordinary — far more energetic than I expected. Over 10,000 views, 1,000+ likes, 190 shares and 189 comments (and counting). The comments were particularly fulsome and encouraged me to be more honest in the future. For now, in case you missed it:

Monday — on what would have been my five-year anniversary at Gartner — I left to join Dentsu Aegis Network. It was a good span at a well-run company doing God’s (technical) work. It was simply time.

When I was a management consultant, I couldn’t describe what I did. Not to my parents, not to strangers. Not in a way that convinced them I had a real occupation, and maybe I didn’t.

Try describing what an “industry analyst” does: “Research, writing, 30-minute consulting engagements.”

“But you can’t solve any problem in 30 minutes.”

“Just watch me.”

“You don’t know anything about the company.”

“I know something.”

“But — but — but –”

The assumption most people make is that marketing problems are unique. Perfect knowledge of the context, the company, its tech stack is required to construct a solution. This assumption is false.

Marketing problems are not unique. There are just a few of them, with variations. It’s the solutions that get complicated.

Anyone can make good progress on a problem in 30 minutes, but they have to really know the problem. Most of the time, they don’t.

My particular areas of coverage were ad campaign measurement, including multitouch attribution (MTA) and marketing mix modeling (MMM); other marketing analytics; and programmatic advertising, particularly data management platforms (DMP) and demand-side platforms (DSP).

So most of my inquiries took the form of: “How can I measure the impact of my ad campaigns?” … or, “How do I organize my campaign data for targeting and measurement?”

Those were the real questions. See what I mean: Not unique. Any marketer could (should) ask them.

The trouble is, the way they are originally phrased in the meeting invitation makes them sound like some fragment of the Dead Sea Scrolls. “We’re in the process of spinning up a 360 view of the customer and we need to find a universal ID we can map to our CRM and MCCM but the customer files are all sitting in the data warehouse we think …”

Real question: “How do I organize my customer data?”

There is no 360 view of the customer. Give it up. Marketing profiles are by nature incomplete. They should be. We’re not building an encyclopedia. If we did, we wouldn’t have time to read it. We need only one piece of information: the right one.

  • She really wants a pair of red shoes now. She’s not price sensitive.
  • He secretly wants to test drive that Tesla but has to convince his boyfriend it’s safe.
  • The only thing he really loves are Bernese mountain dogs.

These are far from 360 but are perfect for marketing.

Most clients of analyst firms do not use the service enough. They mistake it for a Delphic Oracle that is sometimes wrong, rather than a research tap that is often right. Ask a question in advance. Most analysts — like me — spend time preparing for the call. This is called research-on-demand.

Sometimes I think research is my only real talent. And then I remember that I am also a dog trainer. And then — then I realize I am very bad at that.

Vendors sometimes believe analyst firms do P.R. for them or they can pay their way onto a Wave or Magic Quadrant. I have been deep inside the machine, my friends. This belief is not true.

There’s not even much of a benefit to being a client, unless you want to use the service as a service, like other clients. It can help you craft your story, improve your pitch, not sound like everyone else; anticipate end user needs, murmurs in the market; help you emerge from the bubble wrap of hermetic V.C.-and-conference-speak, which seems — from what I can see — to be much better at building brilliant solutions to invisible problems than at finding problems that need to be solved.

Most marketers are not advanced. The real disease we all have is an inability to admit we have no idea what real gangstas are talking about.

“I just wish I were more technical,” a very well-regarded ad tech analyst told me recently.

“We all do,” I said. And it’s true.

Our agenda this year should be to admit what we don’t know, not promote what we do. Don’t nod along if the point is obscure. The talker is faking it anyway. Nobody knows the difference between A.I. and machine learning or why they’re conflated or how to get to a local maxima or why you need XG Boost today when a simple gradient descent algorithm worked yesterday … it’s okay.

If you knew it all you’d be dead, right? There is perfect knowledge in silence. It takes a lifetime to be a good mystic too.

Most vendor briefings are just not very good. They are too long. They wander from the point. From too many it is impossible — and I say this as an intelligent insider who knows a thing or two about your market — completely impossible to say what the product does, exactly.

You don’t do everything. You don’t sell to everyone. Pick your fight. The best predictor I know of a doomed start-up is not a weak board or a psychotic founder, it is a polished pitch deck.

The best start-ups I’ve met are modest and clear, friendly and fairly honest. They do not spend five slides telling me that Millennials crave experience and mobile is the new Web. They do not say the DMP or email or apps are dead. They do not say anything is dead or that everyone else “just doesn’t get it.”

Above all, they do not say they have no competitors.

Think about it, people. Who has no competitors? Who? I’ll tell you. People who are in a market that doesn’t exist. Delusionals. Hallucinators.

Use 12 slides. Don’t have any set-up. Tell me what you do. Name your three competitors. Say what about your tech is proprietary, original, or interesting. Explain the things you’re going to improve. Describe your funding. I liked a demo, if there’s time.

Maybe once or twice a year, I encountered a new company that seemed to me like a winner. I’m trying to think why. They had very little spin and a lot of technical detail, but not as hype. They seemed to be engineers talking to a blogger rather than sales people who had read some white papers. They were somewhat awkward, informal. And their products all emerged from a technical discovery – a thesis, an experimental approach, an application of some new platform to an persistent problem.

A problem like: “How do I reactivate dormant customers?” or “How can I link records faster?” or “What is the real LTV?”

And they all wanted to learn. What have you heard? Is this thing needed? Are we lost in emotion?

Good products take off fast. They surprise their creators with momentum.

If I can indulge in a reflection, I also learned some things about myself. It is important to manage burnout, whatever your age. Working life is a marathon. I’m not talking about getting enough sleep or eating right or exercising; you know that. I mean in the long run, over years, managing a build-up of stress that tears you down from the inside until — one day — you realize you simply can not make another phone call, write another document.

The only way to recover from this situation is to take time off work. As much time as you can afford.

But don’t let yourself get there. You can avoid it. Take vacations. Cultivate a hobby that has nothing to do with your work. Get a Bernese mountain dog. Get married. Hike alone.

Don’t fight fear. Live bravely. Smile more. What we forget about the Hero’s Journey is that it does not change the world. That is not the point.

The journey changes us. We are our only real problem.

Martin Kihn

Call me Marty. And hello.

I am an author, digital marketer, ad agency alum, former management consultant, and programmatic advertising pundit. I was born in Zambia and moved too often, ending up in Michigan. I consider myself to be a Midwestern Gotham-manque. After college, I stormed NYC and was a fact-checker at Spy, reporter for Forbes, feature writer for pubs like New York, GQ and the New York Times.

My best-known writing job was as Head Writer for a program called Pop-Up Video on MTV Networks. We were nominated for a Daytime Emmy, and my mother has a picture of me and then-Mayor Rudy Giuliani. His tan is smoother.

After business school, I became a management consultant and published an irreverent memoir of that experience called House of Lies. This book later formed the basis of a Showtime series of the same name starring Don Cheadle as a highly stylized version of myself named Marty Kaan. It ran from 2012-2016.

You can see me on the right side of the stage here with the stars and producers of the show:

houseoflies-600 x 400

My second memoir was a send-up of self-help books called A$$hole: How I Got Rich and Happy By Not Giving a Damn About Anyone & How You Can Too. It was in development for years at Warner Bros. but never quite happened. However, it was a genuine best-seller in Germany, where I hope they caught the irony.

My most recent memoir is Bad Dog (A Love Story). Altogether less snarky and more emotionally satisfying — at least, to me — this book told the story of my beloved Bernese mountain dog, Hola, who taught me how to be a sober man by training me for our Canine Good Citizen certification. It too was in development (at NBC) for a time.

Professionally, I’ve worked on the analytics and measurement side of the digital advertising business since 2004, at agencies in NYC and Minneapolis. Currently, I’m a Research V.P. at Gartner, covering advertising technology and digital marketing analytics.

Much of my professional writing, which has won internal prizes (twice), sits behind the Gartner firewall. I’m also one of the firm’s most-read bloggers, and I contribute pieces from time to time to various worthy sites like the mighty AdExchanger.

My passions in life are modern ballet, particularly NYC Ballet and this person, and Bernese mountain dogs. I live in Katonah, NY with my wife, the singer-songwriter Julia Douglass, our BMD Jordan, and this rather dominating personality, Jerry:

jerry