Why Would You Use Leaky Buckets to Handle Your Precious Customer Data?
In today's complex martech environment, customer data never stands still.
From the time it enters your ecosystem (for instance, via a POS transaction, or a membership drive) to the time it gets matched to an active marketing campaign (either directly or via an onboarding platform) and eventually comes back for measurement, your customer data goes through a staggering number of platforms and point solutions.
One estimate a few years ago put that number at 91 different martech tools for the average organization. If that seems like a lot, consider that the Martech supergraphic (a comprehensive directory of all martech solutions on the market) started out with 150 vendors 10 years ago and now has over 8,000 listed companies. There's clearly no shortage of options.
But every time your data is handed off from one tool to the next, there's a good chance that you're losing something important. Each tool comes with its own idiosyncratic data rules and formats, and its own refresh cadence as well. With every hand-off, the data gets out of sync and errors creep in.
Do you remember the leaky bucket relay game you played as a child? You would fill up a leaky bucket with water at one end of the field, run as fast as possible to empty whatever was left in the bucket into a container at the other end of the field, and run back to hand that leaky bucket to the next member of your team.
That was a lot of fun, and a sure way to spill lots of water all over yourself on a hot summer day, but it's not a game you want to play with your customer data.
How bad is customer data decay?
Every year, 35 million U.S. consumers change their phone number. More than two million change their legal name. We measured that it takes less than two years for more than half of the data in a CRM database to go obsolete, and that was before millions of people hit the road (and changed their mailing address) during the COVID-19 pandemic.
If you don’t take measures to fend off data decay and keep your first-party data as current as possible, you’re essentially spending money on customers that don’t exist and ruining your marketing ROI. You might as well exit data-driven marketing altogether and buy 30-second ads during the Super Bowl.
How about data leakage?
While data slowly decays within each of your martech tools, the unfettered hand-off of that data from platform to platform produces serious data leaks too.
A company came to us recently because it wasn't seeing the results it had expected. Its identity provider could match only 75% of the organization's customer records, so 25% were left unmatched. Its CRM then fed those records to a customer data platform (CDP) where 15% failed to match. In its onboarding platform, another 40% was lost to poor matching of offline to online data points, and its DMP exacerbated the problem by another 45%. Altogether, only 20% of the organization's original dataset made it through unscathed.
It started off with 100 million consumers, and after this painful game of leaky buckets ended up with 20 million. You'd be upset too, right? And that was with only four platforms in its marketing stack—imagine the loss in data fidelity at a company that boasts 91 martech tools.
Is data duplication a problem too?
With leaks and decay come duplicates. If every tool in your marketing stack has a different view of identity and no tie-back to a reliable source of truth, can you tell if Archibald Buttle really is Archibald Buttle the cobbler, or Archibald Tuttle the heating engineer? In Brazil, Terry Gilliam’s 1985 cult-film classic, a hapless fly jams a teleprinter just as it’s printing out an arrest warrant for Archibald Tuttle, and Buttle ends up arrested instead (and killed during interrogation). Small details matter.
Data duplication is one of those problems that’s hiding in plain sight: We worked with a major retailer recently to help it raise the quality of its CRM data—and therefore the efficiency of all its data-driven campaigns—and found that out of 300 million records in the organization's database, 100 million were duplicates. That’s not out of the ordinary, by the way. IBM places the overall rate of data duplication around 20%-40%, and that’s generally what we’ve encountered too in our client engagements across most industries.
How to fix the problem?
In a recent report commissioned by Neustar, Forrester estimated that barely 28% of marketers are confident that they have the tools they need to address their organization’s current data management challenges.
Every single tool has room for improvement, for sure. But the biggest breakthrough comes when companies start to trim their marketing stack. When there are too many data hand-offs between siloed platforms, there are simply too many points of failure. Consolidation is the operative word.
It’s also crucial for the remaining data platforms to espouse a unified identity framework. You don’t want your POS system to ID customers by their telephone number; your CRM system by their age, gender, and zip code; and your onboarding platform by their email address. And you definitely don’t want two dozen different IDs floating around for the same person. Think I’m exaggerating? We did a data health assessment recently for a client and found that it used 46 different identifiers to recognize its customers. The company was as horrified as we were.
Naturally, for unified identity to work at its best, you need a reliable partner that can help you integrate all the platforms that remain into a cohesive whole, and serve as a truth-set every step of the way.
It’s as if you asked every kid in the leaky bucket game to hold their hand over the hole in the bucket and walk real slow. It’s not quite as fun as the original game, but if there's something you don't want to do with your customer data, it's spilling it all over the playing field.
Ready to stop playing games with your customer data? Please contact us and we’ll show you how to use unified identity to consolidate your marketing stack.