At 97th Floor we love to talk shop. We also love to talk about pop culture, and you better believe that television tropes are frequent points of discussion. You may have heard of the classic trope known as, A.I. is a crapshoot. It’s the scenario where a supposed perfectly created artificial intelligence rises up, and ultimately leads to humanity's demise — or at the very least ruins someone’s day.

This isn’t a new concept. In fact, it features (in varying degrees) in some of the worlds most well known movies and television franchises: Terminator, The Matrix, Avengers: Age of Ultron, 2001: A Space Odyssey, TRON, Logan’s Run, Battlestar Galactica, Dr. Who, and the list continues on until you get to the genteel Pixar classic, WALL-E (I am of course referring to the evil co-pilot robot, not the cute one that saves the world).

If pop culture is any indication, it’s pretty clear that we, as a collective whole, have trust issues when it comes to cold, soulless data. But if that’s the case, then why are marketers so keen on putting the numbers in charge?

Uniquely Data Driven (Like Everyone Else)!

Now, don’t get me wrong; I love data. But I have to love data. A digital marketer literally cannot function without data. So when I see “data-driven” on a resume or across an agency’s site, it’s almost like they’re saying nothing at all.

It’s like seeing a pizzeria advertising that they use dough on all of their pizzas. I mean, sure, it’s better than the alternative (although dough-less pizza would be a big hit with the keto crowd). It’s probably just not something that needs to be advertised. It’s a pizzeria; even without being told, we’re already pretty sure that dough is part of the equation.Data and modern marketing are inseparable. So stop bragging about it.

Data Should Never Dictate

Additionally, I believe there’s a growing flaw in the data-driven mindset, namely that data becomes a dictator in the strategy. But handing over the reigns to analytics programs is exactly how LA got cremated inTerminator, and doing the same thing with marketing strategies is only slightly less careless.

If you’ve ever read a terribly unhelpful blog post, you could probably trace it back to some SEO who saw a valuable long-tail keyword, and then threw together some heavily optimized (but ultimately worthless) dumpster fire of content designed to capture it. Or, if you’ve ever been accosted over email relentlessly asking for a purchase with varying price discounts, it’s probably because the data said that users make purchase after 14.2 emails, which led a marketing automation specialist to believe that hitting you up on a daily basis would make you convert faster.

And yes, obviously analytical insights are invaluable. And strategies built exclusively on data can even work sometimes. But the returns will seldom stack up to what you can get with a flesh-and-blood marketer calling the shots. The point is that data doesn’t run marketing strategy for humans. Humans make campaigns for humans, and we need to keep it that way.

In other words, you never put the robot in charge of the nukes, and you don’t let a spreadsheet do your marketing for you.

Data Is the Beginning of Marketing Strategy, Not The End

But PJ, you’re in digital marketing. Are you saying you ‘don’t use data?’

No. I’m not saying that I don’t use data. Quite the contrary, we need data. We rely on it — always have, and always will. In fact, let me tell you a story:

In the first version of competitive research I ran at 97th Floor, we collected 43 metrics for each result on the first page of Google, plus the page we were trying to rank.

Don’t worry, I’ll do the math for you: That’s 473 different cells being manually populated. And when I say manually, I really do mean manually. We used 10 different tools to get everything we needed. Metrics ranged from word count, to load time, to followed backlinks, to Google+ shares (do you feel old yet?). A single SERP would take someone 1.5 – 3 hours. Let’s just say I got through a lot of podcasts during this method of competitive research.

And while time consuming, this method of data mining was therapeutic, restful, even invigorating. The process of gathering data by hand was lengthy, but every hour in the trenches saved weeks of bad strategy.

A lot has changed since then at 97th Floor, we’ve scaled that research process into our suite of proprietary software. Now of course we have our competitive research tool (more on that to come later), which lifts all the metrics we need through the glorious use of APIs and some custom formulas. But the core truth, that our data helped us do a better job, remains.

So yeah, we like data. We respect data. We use it and abuse it. But data, all by itself, does not move the needle — like at all. It’s the action you take based on the data that makes all the difference.

I’d be lying if I said that we’re not proud of our new set of proprietary tools that have essentially taken our drawn out processes, and condensed them into something simple, effective, and efficient. But the reason we love these tools so much is because of the time they save us on the front end of a campaign. If we can condense the first week’s worth of data gathering into an afternoon of well coordinated data pulls and analysis, we’re that much further ahead. But we’re certainly not done.

Has Data Ever Saved the Day?

Not all robots are bad. Heck, R2D2 certainly picked up more than his share of the slack. And sure, most of the time data does the heavy lifting.

I remember a relatively recent situation where a newly onboarded client approached us with a keyword they had been struggling with for years. They had a well optimized page for it, the page had many links going to it. On paper they should have been doing better for this keyword.

Once upon a time, that page had been in an above-the-fold position for its SERP. Then it had begun to drop. The client added more copy to the page to counter the dropping ranking. But more copy caused it to drop even further, which suggested to the client that they had to add even more copy. This game of cat and mouse went on until they were off the first page entirely. And that’s where they were when 97th Floor was brought in.

After assessing the situation, we threw the page into our proprietary competitive research tool. Once the data was finished crunching, the answer was obvious. They had vastly over-optimized the page!

When looking at the top 10 results on the front page of Google, the SERP was seeing an average word count of 1,983. But the page our client has built had… any guesses?

17,744 words.

Not characters, words. Oh and the page was using the exact-match keyword more than twice the amount of the average on the SERP.


The data called out a problem. It helped save the day. But the data didn’t come up with the solution.

We got to work crafting a strategy involving everything from keyword usage, to semantic analysis, strategic internal links to alleviate cross page cannibalization, and strategically cutting and reworking sections of the page. Essentially, we deoptimized the page. After the above plan was executed with the client, we saw fruit — jumping from page 2, to spot 2 in less than a week.

I love this story because:

1) The tool saved us a bunch of time upfront on research and on the backend by ensuring we weren’t wasting time reoptimizing what was already too optimized.

2) The SERP data lead us deeper down a very productive rabbit hole; cannibalization, which upon more research, yielded some very actionable tactics.

3) Possibly my favorite, we didn’t have to waste any time and energy on backlinks.

Data Doesn’t Move the Needle; You do

The above story and hundreds before it have happened at 97th Floor. While we can’t function without data, we certainly can’t function with numbers alone. That’s why it’s called digital marketing, not digital analyzing.

That’s why I love this industry. We’re a fun breed of humans that crave and cultivate good data, and then create something bold and beautiful — with or without the data’s support.

I cringe a little when I hear people say, “well, this is what the data says we should do.” I’m sorry did I hear you just imply that this spreadsheet told you to increase your bid strategy? Of course not. What you meant to say was, “I believe we should increase our bid strategy.”

Own that strategy, don’t let data own you.

Don’t be handcuffed by data; have the courage to look beyond surface-level data inferences, and use your own intelligent, problem solving, human brain to figure out the rest. Because, when all is said and done, your clients aren’t interested in hiring a spreadsheet (or even R2D2).