The Cost of Bad Data

What’s the better investment… buying a luxury you want, or fixing a problem you have? 

By far, fixing today’s problems will pay off much bigger and sooner than splurging on tomorrow’s dreams. 

Photo Credit: Luke Chesser/Unsplash

It’s the concept of addition by subtraction. Remove the right negatives and you’ll see a better result than if you added new products or features. Simply put… more isn’t necessarily better. 

However, that concept runs counter to the way the marketing industry views consumer data, where scale is rivaled only by inaccuracy. This is a huge risk, given the importance of data in today’s world. Data is the new oil. It drives AI, the revolutionary technology that stands to have more global impact than the Industrial Revolution. 

Yet all marketing data today is built on massive error. The data compiled is based on guesswork and inferences. Somebody bought something, somewhere, and that data is then interpreted to represent a person who’s maybe a certain gender, of a certain age, in a certain location, who likes certain things. 

Device ID data, such as third-party cookies, are mostly noise with very little signal. A cookie or an IP address isn’t a human being. They don’t make choices on what to watch or what to buy. They don’t have kids, pets, or homes. They’re just numbers in a database, from which we try to make educated guesses. 

But there are a whole lot of them. 

And since the historical answer to bad data is usually more data, marketers continue to buy up vast volumes of consumer data, hoping to build as large a net as possible. The bigger the net, the thinking goes, the more likely you are to catch something valuable. 

That’s the equivalent of spending on new luxuries rather than on fixing existing problems… the data they already have. What’s more, these nets have pretty significant holes far larger than you may realize. And these holes are costing you money.

A Shot In The Dark

We conducted a study last year that measured the accuracy of postal addresses matched to hashed email addresses in the data provided by a range of providers. We found the average accuracy score was 51%, with a low of 32% and a high of 63%.

What’s more, according to our analysis based on the data scoring we’ve conducted to date: 

• 39% of budgets targeting specific genders are misspent. Even if you just guessed, you’d be right about as much of the time as spending money on data that’s only half accurate. 

• 83% of age segments are inaccurate. 

• 70% of campaign impressions supposedly served to Hispanic audiences were not being delivered, while 73% of ads targeting Black audiences are also wasted.

These aren’t rounding errors. They’re massive holes, both in the accuracy of the data and the effectiveness of the money spent to acquire it. 

This highlights not only the importance of choosing the right data provider but also the need to filter the data received from them more granularly, so the money spent on these efforts is more effective. 

To be clear, the providers of this data are not at fault here. The traditional business model for data is simply outdated. In the old broadcast world, networks built large audiences and then sold those audiences to advertisers. But in today’s world, using inaccurate data to make business decisions, like ad targeting or personalized recommendations, results in errors and waste. This results in subpar results and ultimately limited profitability.

From Data To Dollars

According to MIT, bad data can cost organizations as much as 25 percent of total revenue. And IBM famously estimated that bad data costs U.S. businesses $3.1 trillion a year. 

Meanwhile, improving the quality of the data used to communicate with customers has been found to both save money and increase revenues. A Truthset/CIMM study found that a typical consumer packaged goods campaign could boost its returns from $1.08 to $1.54 per dollar invested in data by improving the quality of identity data and match rates. 

At Truthset, we’ve found that improving the precision and profitability of data-driven marketing by just 10-30%, through the process of identifying and removing errors, can yield over $100B in value. 

How? With better data, brands can filter out inaccurate records in an audience file before targeting to drive better relevance and results. Publishers can improve the fidelity of their audience data and expand their identity keys to increase match rates. Measurement companies can align to a common objective accuracy standard for exposed audiences.

The old spray-and-pray marketing paradigm is over. The world is moving to a more targeted landscape where customers expect personalized, relevant ads catered to their individual needs. Yet today’s audiences are fragmented across multiple platforms, content, and sales channels. That makes sending the right message to a smaller, more targeted audience far more impactful (and necessary) than a generic message designed for all. 

The future is about precision. By identifying and removing inaccurate data before you send a single email, you can immediately lift your campaign targeting and performance KPIs and more importantly, reach more relevant audiences. 

So before you spend more money buying more data in hopes of reaching more customers, take a moment to consider the advantages of instead investing in improving the data you have. It’s no different than fixing your squeaky floorboards before paying for that new addition to the house. 

Removing negatives from your life first is always more efficient than adding perceived positives.