Weeding Out Bad B2B Data Is Key to ABM Success
Featuring RK Maniyani, CTO, True Influence
The incredible volume of user data now available to B2B Marketers has opened doors to new levels of audience segmentation, targeting and content personalization.
The amount of data being created annually was estimated to grow by more than 50 percent last year, according to predictive analysis company Radius. And that rate is projected to keep growing as B2B prospects use even more devices and channels to gather information for their purchase decisions.
But with all that data comes increased risk of poor data quality, the nemesis of efficient Account-Based Marketing (ABM).
Poor data quality not only hurts performance of current B2B campaigns, it also creates faulty analysis that undermines future efforts in the increasingly “right-time” model of automated and quick-response B2B Marketing campaigns. And an empty pipeline means decreased revenue, which is good for no one.
Just last year, Gartner cited the impact of poor data quality at $9.7 million a year per organization, up from an estimate of $8.8 million in 2015. And the analyst firm expects the problem to escalate as more data sources and complexity enter the mix. Not all bad data is directly associated with B2B Marketing, of course, but surveys show that 80 percent or more of marketing organizations have problems with bad data, and Discover.Org reports that old or incomplete data is the main source of grief for Chief Marketing Officers.
Many analysts, including Gartner, are calling for 2018 to be the year that B2B Marketers and data teams finally elevate dealing with the root issue of bad data, with a focus in investing in modern data quality tools.
But in my opinion, the persistence of data quality problems indicates that many organizations still need to optomize their core practices and policies for basic data onboarding and management. This position is backed up, I think, by Radius’ findings that a simple lack of data management skills is a top obstacle to improving data quality.
How should an organization begin to improve its data quality in 2018? The first step is to clearly understand how problems can be spotted in key data channels and addressed before they undermine full-scale marketing campaigns.
Bad Contact Data
Even in the era of sophisticated behavioral targeting, much of the conversation surrounding bad B2B Marketing data focuses on the most basic unit of customer data, contact information.
By its nature, contact information ages rapidly. Research shows that in a three-month period, about 8 percent of contacts in the average CRM become unreachable due to job changes, relocation or other factors. That degradation alone will completely undercut campaign KPIs, if not monitored and corrected. In fact, Sirius Decisions suggests that a single bad record in your Customer Relationship Management (CRM) system costs about $100.
In addition to aging, contact data can be rendered useless through data-entry errors and poor data standards / integration rules between sources, both internal and external.
Obviously, monitoring and scrubbing contact data is essential for B2B Marketing success. Aside from the basics of purging non-responsive emails and duplicate records (Hubspot offers a nice primer on these tactics), you should run periodic tests against your database for readily identifiable errors, such as location or company info that contradicts IP and other location tracking data. This is especially critical in ABM, since the Marketing team is building interest-level profiles for overall accounts, not individuals.
And, based on the revenue potential of each prospect in your CRM, phone verification and enrichment of contacts can be a worthwhile investment. Reputable third-party lead sources, such as True Influence’s InsightBASE service, verify the contact information they add to your pipeline – obviously, the ability to successfully reach a new lead is essential. Running similar phone bank programs against sample pools of older contacts will give you a good snapshot of the quality of your CRM data, as well as possibly “warming up” some older contacts for new messaging efforts.
Bad Targeting Data
You can also encounter serious issues with the behavioral data essential to identifying active demand, which is critical to effective B2B Marketing.
In addition to systems that track user behavior on you own Webs sites and emails, successful ABM campaigns require incorporation of third-party data, such as the Intent Signal Monitoring data available in our InsightBASE product, to get a complete picture of prospects’ purchasing profile. (No less an authority than SiriusDecisions considers this an essential asset for ABM).
The sheer volume and complexity of these data sources demand that your team perform quality checks as you onboard data into your systems, prior to executing any large-scale campaigns against it.
Obvious anomalies in reported behavior patterns are the primary red flag to watch for while onboarding behavioral data. A company’s interest in a topic, as expressed in Intent Signal Data, is not going to explode from no searches to 100,000 searches in just a month. Every email you send to an account is not going to be viewed.
Not every behavioral anomaly is going to be as obvious as the examples above, of course. Your data team should partner with business managers (more on that later) to identify a short list of weird patterns to identify and suppress before data is onboarded. And this process needs to be scheduled and executed promptly, since fresh data is essential to active demand marketing.
And, as I said earlier, you need to understand sourcing and capture methods – the color and the shape of the data, if you will – as you evaluate data quality and patterns that just can’t be right.
The primary sources of noise that can create errors in targeting data are:
Incorrectly Tagged Data
If data is being assigned to the wrong account or geography, you can get a wildly inaccurate picture of the actual level interest in a topic being exhibited by accounts in your ABM strategy. Mobile users, particularly those that travel extensively for work, can create confusing matching patterns – is this person just working out of a coffee shop on the West Coast, or are they not actually an employee of the account you have them tagged for?
A key issue here is whether the data is being collected via IP or other location-based technologies, or if it is gathered via device-based cookies (which creates a whole new set of issues to consider).
Other tagging issues include acquisition channel and other interactions than can seriously injure low-funnel campaigns down the road.
If even Google is struggling control bots’ impact on data quality, it’s an issue for everyone, including B2B Marketer and data teams.
Our own Craig Weiss recently shared some interesting insights on the potential impact of increasingly intelligent bots on Intent Signal and other behavioral data, but today the main impact of bots is unrealistic email and advertising performance. You simply are not going to see 50 percent CTRs on even the most artfully targeted campaign.
Again, your team needs to identify reasonable performance metrics, based on verified past activity, and suppress data that simply can’t be accurate. Deeper investigation will likely identify the root issue of the problem later, but your first concern should be to not onboard obviously flawed data that will end up wasting Marketing resources on ill-designed and poor-performing campaigns.
Data Quality Is an Ongoing Initiative
As you’ve probably gathered from this article, I believe that ongoing collaboration between your data team and business managers is essential to ongoing data quality efforts. I’ve written earlier that I believe data quality is ultimately best owned by data scientists within the technology organization, and this is still the case.
But business managers are the best sources for the expected range of patterns your organization should set as the baseline before onboarding new data. And they also are a first line of defense in reporting new anomalies as they pop up in marketing campaigns. A cross-disciplinary data quality team should meet regularly to review campaign performance, to check for obvious anomalies, and to refine your data quality standards. Business managers will be an invaluable resource for developing data quality tests, not only for new data but also against sample sets in your production data stores to ensure they are still valid.
Eventually, this data quality team can also collaborate to find more complicated – and potentially more lucrative – patterns in user data. That’s where advanced technologies such as Machine Learning can help fully exploit the enormous potential of Big Data to drive B2B Marketing.
But to tackle the pernicious issue of bad or outdated data, the first, most important step is to impose a discipline of identifying anomalies before onboarding new data and testing samples of production data to ensure your ABM campaigns hit their targets.