Five Words of Warning for Corporate America: How Big Data Lost Us the Vietnam War

In Big Data by Nick JaworskiLeave a Comment

Five Words of Warning for Corporate America: How Big Data Lost Us the Vietnam War

Despite being involved in a number of subsequent wars, the Vietnam War remains one of the most vivid and painful memories of a war waged by Americans abroad. After victory in WWI and WWII, this was America’s biggest defeat, causing extensive rifts within the country itself.

While there are certainly many factors that led to the US losing this war, big data was one of the biggest. Within this monumental loss lies a warning to all the senior executives that now prescribe to the fetishization of data.

What lost America the war is what is also destroying companies today. We’ll look at the 5 ways big data lost us the war, but, first, a little history about one very important man.

The Architect of the Vietnam War

Robert McNamara is known as “The Architect of the Vietnam War.” As President Kennedy’s Secretary of Defense, he was in charge of all overarching strategy for the conflict.

And he was obsessed with data.

McNamara graduated from Berkeley with a BA in economics and then Harvard with an MBA. Afterwards, he entered the armed forces and was most notable in his role in WWII at the Office of Statistical Control, an office that used data to solve major problems of war such as inventory management and supply chain operations.

McNamara was so successful in his role with the Office of Statistical Control, they were reported to save 3.6 billion dollars in armament procurement in one year alone.

After the war, he was quickly scooped up by the, at the time, troubled Ford Motor Company. Using data within every decision, he quickly moved up to President of the company before being chosen to join Kennedy’s cabinet as Secretary of Defense.

Throughout his life and career, McNamara was a supremely rational man, one who believed good decisions could only be made from data. He also strongly believed there was a right answer for every situation, you just needed enough data to find it.

Does any of this sound familiar to those of you in today’s corporate world?

As Secretary of Defense, he applied this rationale to every decision involving the war. In a war that led to the death of over 58,000 American soldiers, we’ll see how big data aided in their deaths and what warnings this has for any corporate executive that’s overly zealous about its use as well.

  1. Data Lacks Context

This is probably the biggest problem with the over-reliance on data. McNamara had tunnel vision. He was focused on only one thing – body count.

Body count was the determining metric for everything related to the war. The entire goal was to kill so many of the other side that they either surrendered or were forced to the negotiating table.

Yet, war is far from so simple. Take D-day as an excellent counter example. The number of Allied soldiers killed on the first day of the invasion were over 10,000. Axis forces only lost 1,000. Even though Allied deaths far exceeded those of Axis soldiers, the conclusion of this battle was a tremendous victory for the Allies due to ground gained, psychological blow to enemy forces, and increased strategic positioning.

It is often the context of actions that determines their true outcomes. Numbers are simply one part of that story, and sometimes a very small part.

I’ll give a couple examples from my time at Disney to further illustrate these points. I once had the Regional Vice President contact me asking me to seriously review and probably dismiss one of my teachers due to his incredibly low parent satisfaction scores.

However, the fact of the matter was that he was the 7th teacher in that classroom in less than 6 months. Disney had struggled extensively with teacher retention in its Chinese schools and parents were also bound by a contract. Since they paid up front, they were stuck with us until the end of their contract.

The teacher in the classroom at the time was excellent, certainly one the best out of the previous six. I actually promoted him about a year after this.

The parents were extremely upset about their situation and all the previous teachers. It had nothing to do with him. But all this RVP saw was the data in a spreadsheet. Had we acted on that pressure from the top, we’d have had to move on to our 8th teacher in 6 months. Knowing the story on the ground, that wasn’t a decision I was about to make.

  1. People Fudge the Numbers

As one Vietnam War general recounted relating the oh-so-important metric of enemy body count, “They {the numbers} were grossly exaggerated by many units primarily because of the incredible interest shown by people like McNamara.”

Data is only useful if it retains integrity. There will always be errors in the data. Using a marketing example, Facebook generally reports a 30 percent higher number of web traffic from the site than Google Analytics does. Facebook has a vested interest in making you think it’s driving more traffic to your site than it probably is.

In the same way, people want to look good to the bosses. When I worked at Disney, the practice of padding reports was called “fluffing,” and most often occurred during projections and forecasting. Managers would always inflate their sales goals because to do otherwise was to make oneself look like they weren’t ambitious enough.

In another example from Disney, for nearly the first three and a half years of the company, we reported that over 90 percent of our students got 100 percent or higher on our internal exams. Have you ever heard of a school where 90 percent of the children are not just getting “A”s, but getting flawless scores for years? Me neither.

Obviously, even casual observation showed the tests were incredibly easy, but, moreover, teachers were evaluated on student performance. So nearly all of them would “help” students struggling on the test, leading to miraculously flawless scores time after time.

This led to the even greater problem of assuming our curriculum and teaching methodologies were much more effective than they really were. When changes for improvement were suggested, the data was used as a reason to not make any, even though all of us actually interacting with the students knew that there were huge deficits.

If people are more focused on the numbers than on achieving a goal or delivering an outcome, they’ll get lost counting sand rather than actually crossing the desert.

  1. Formulas Aren’t a Strategy

For McNamara, is was all about finding the winning formula. How many troops, guns, or Apache helicopters would he need to move to position X to increase enemy body count and win the war?

Most of us looking at this would find such a question ridiculous. Of course, winning a war is far more than a mathematical formula. McNamara, and his obsession with data-driven decision making, didn’t see it that way.

I get this all the time in the marketing world. Company A wants to give my company X amount of dollars. In return, they want 2, 3, or 4X amount of dollars back by such and such date. If marketing were so easy as to simply apply a series of actions to get X amount of return, we’d all be billionaires.

I’d say, “as anyone who runs a business knows, this isn’t the case,” but that’d be inaccurate as most people asking my company to do this actually do run businesses.

We can get more granular, and we often do these breakdowns with clients. You need X amount of revenue per month. You have a lead-to-sale conversion rate of 60%, an on-page conversion rate of 10% for visitor to lead, and so you need to send Y thousands of people per month to your site to achieve your revenue goals.

Again, we all know it doesn’t really work out half so neatly. There are a million unquantifiable factors in those equations such as mood of the customer, skills of the sales team, current market demand for product/service, quality of website visitors, and on and on.

The math is certainly an excellent starting point and absolutely necessary to making strategic investments in your company’s growth, but finding a formula will never be more than partially predictive. There are simply too many other factors involved that you can’t control.

What this leads to is the exact same problem McNamara had. Rather than searching for another solution, they keep trying to fix the formula. In the marketing world, this often means bouncing from agency to agency or Marketing Director to Marketing Director.

The theory goes, if agency or director X can’t get the formula right, there must be someone out there who can. The reality is that it takes one very strong team to constantly work on a problem together and keep iterating until success is achieved. Each time you simply try to change the people in the formula, you have to start over. This causes huge lags in ability to achieve breakout success.

  1. It Can Be Misinterpreted

Misinterpreting data can be disastrous. In fact, this was another big reason America lost the war. According to the data and McNamara’s formula, we were winning. The Viet Cong’s body count was regularly higher than ours. Compared to the paltry 50,000+ America lost, the Vietnamese government has reported over 1 million Vietnamese killed in the conflict.

Clearly, we were on the verge of winning and just needed to keep at it until the Viet Cong succumbed to our rational formulas, as logic would dictate. This continued for going on 20 years.

Just because you have data doesn’t mean your analysis of it is accurate. Going back to our Disney example of all the brilliant children passing with 100 percent, the assumption of the corporate office and curriculum team was that the curriculum and teachers were so good, all the kids were learning incredibly well.

However, just having simple conversations with the children around previous lesson material would quickly show you that wasn’t the case.

  1. Data Paralysis Is a Thing, a Big Thing

Data can both paralyze organizations by making them fear to act if they don’t have enough of it, and paralyze them by spending too much time in its collection, analysis, and interpretation.

Generals on the ground, actually in the conflict, weren’t being asked about strategic direction. They were being asked for numbers. In a war, split-second decision making is often crucial. By the time one has data sent back to the Pentagon, scrutinized by analysts, and then interpreted to determine a course of action, everyone on the front lines is already dead.

And imagine the amount of time spent in creating and reporting on all the metrics McNamara put in place. How could time and resources have been directed to strategy and execution without such overly thorough KPIs, oops, I mean reporting metrics, in place.

In my time at Disney again, reports were a big thing. Under some higher ups, we were reporting sales numbers up to 4 times a day! This was in an environment where 5 was a lot of sales in a day, so it wasn’t like we had huge numbers we were dealing with that might need more frequent reporting.

On top of that, if we didn’t have good numbers in the first 2 hours of the day, we’d have to spend extra time explaining why and then outlining what we were going to do in the next 2 hours to up the numbers. I would not be amiss in saying that we sometimes spent more times reporting on sales than actually working to get them.

The same happens in the digital marketing world. We are flooded with data and we’ll have clients that request weekly, sometimes even daily reporting. Yet, that level of frequency simply has no statistical significance for all but the largest businesses.

If we take the Facebook algorithm for an example, it generally takes it 7 days to optimize any given ad set. So if you’re constantly tweaking things, you’re shooting your campaigns in the foot because Facebook’s machine learning, which is incredibly powerful, is not able to assist.

This also goes back to the formula piece. Marketing is very much about building momentum. The formula you have in place may actually work, but you won’t see it start to work for a sometimes months. Patience is hard in our Wall Street world of quarterly reports, but successful businesses learn how to play the long as well as the short game.

You can always get more data, but execution will always trump strategy. So get enough data to help it inform your decisions, and then act. The opportunity cost of waiting is often much higher than anything to be gained by ever more data.

Remember, It’s Evidence-based, Not Evidence Only

These 5 warnings should ring true and be well heeded by anyone in today’s business world. We don’t want our organizations to suffer the same level of losses America did under McNamara’s guidance.

Big data is vital to our organizations, but it is only a starting point. It should never be the sole source for any decisions. There are simply too many other factors.

The true power of machine learning is to be able to mine vast quantities of data quickly and present correlations a human mind may have taken years to find.

But the human mind’s unique ability is still to make creative and innovative predictions in a highly variable and uncertain world. Our ability to act and execute, rather than just analyze, also currently sets us apart from the machines.

And our people are not formulas or numbers. They should never be treated as such or we all lose.

As a business leader, understand the context, the story behind all the data, and use the genius and talents of your team to make decisions that are evidence-based, but not 100% determined by said evidence.

Photo Credit: martinlouis2212 Flickr via Compfight cc

This article was first published on Integrated Marketing Association.

Nick Jaworski

Nick Jaworski was listed as one of the top 250 social media marketers in the US and owns Circle Social Inc., a strategic digital marketing agency in Indianapolis.

As a digital strategy and data nut, he loves building tactical campaigns across digital assets in order to create online lead and sales funnels. He can regularly be found sharing his knowledge through his writing and conference speaking.

When he's not on social media, he's spending time with his favorite person in the world - his daughter.