7 Strategies to Combat Bad Data Quality in B2B Marketing

A 2019 study by Forrester and Marketing Evolution found that 21% of marketing budgets were wasted as a result of bad data. According to the study, other ways marketing teams were impacted by bad data include inaccurate targeting, loss of customers, reduced productivity, poor customer experience, and inaccurate marketing performance.

Managing bad data is an inevitable challenge for all teams and organizations, regardless of their size or stage. If you mention “bad data quality” to a room of marketers, you can bet they’ll each be thinking of entirely different scenarios representing what “bad data” means to them. That’s because bad data in sales and marketing is often a broad, vaguely discussed topic that can rear its ugly head in a thousand different ways.

In this post, we cover some of the ways bad data creeps in on marketing programs, and the strategies and tactics used to clean and maintain your data over time.

Common Examples of Bad Data

Hello, [Null]!

Incomplete customer profiles can throw a wrench in marketing campaigns or analytics projects. Underreported segmentations, inaccurate analyses, and missed revenue opportunities are some of the unfortunate consequences of “spotty” data records.

Incomplete data records can result from a variety of scenarios, but some common ones include:

  • Multiple customer profiles in different tools
  • Manually filling out profiles
  • Progressive profiling lead forms
  • Incomplete form submissions
  • Unavailable data

Getting the data in a usable state can be time-consuming and tedious, often involving manual spot-checks to fill the gaps.

“I noticed you work at Urjanet…or, at least you did last year?”

As I scrolled through Facebook one day, a uniquely personalized ad caught my attention.

Generate 4x more leads for Urjanet, the ad read.

The problem: I hadn’t worked for Urjanet for seven months.

Personalized marketing campaigns like these start out with great intentions, but their execution can fall short of expectation if the data is outdated.

Another example of when stale data can come back to bite is when your acquisition marketing campaign targets a lead that is already talking to sales. In this scenario, your tools don’t have the most relevant, up-to-date information on your leads, resulting in wasted spend.

United States vs. U.S.A.

Unstandardized data comes in many different forms and can wreak havoc on marketing operations and analytics. Data that isn’t presented in a standardized format can result in potential leads “falling through the cracks” when it comes time for campaign preparation, which can lead to missed sales pipeline and revenue opportunities.

For example, if your customer data platform or marketing automation tool shows one-third of its Account records tagged with the Revenue field as a range of options (e.g. $20-50M), while the rest are represented as an absolute number (e.g. $25,000,000), your campaign audience may miss out on a good chunk of potential candidates.

Merging datasets together, human error, and multiple contributors (like a large sales team) are all reasons why data can look different from one record to another. Some common places unstandardized data can appear include:

  • First-letter capitalization (e.g. mary vs. Mary)
  • Job title (e.g. CEO vs. Chief Executive Officer)
  • Company suffixes (e.g. Inc, LLC, LLP)
  • Revenue and employee count (e.g. picklist of options vs. absolute number)
  • Address fields (e.g. GA vs. Georgia)

Without the right tools and processes, standardizing data can be a tedious effort.

“Are you Angela Sun? Or Angela Sun? Or this third Angela Sun?!”

The repercussions of duplicate data can range from moderate annoyances (like two sales reps realizing they’ve been working the same account because two records for The Home Depot exist in their CRM) to very time-consuming and costly clean-up efforts. On either end of the spectrum, duplicate data is a major time-waster.

A few real-life examples include:

  • You imported a CSV file, but did not specify which field value to match on
  • A native API integration between two software applications doesn’t check for existing records before creating them
  • A sales rep mistakenly creates an Account without checking to see if it already exists
  • A lead fills out a web form with their personal email address, then fills out another web form with their business email address

[[DEPRECATED]] legacy fields

On several occasions I’ve thought, Man, wouldn’t it be great to just start fresh, with a blank, new CRM? CRMs are often where good, clean data goes to die. Multiple stakeholders and data contributors, no CRM owner, mismatched data mapping with other tools, and an unclear strategy are all factors that can create the perfect storm of data chaos.

Multiple fields (for the same thing) or legacy fields that haven’t been properly deprecated can cause enormous confusion, especially if you’ve inherited a CRM or marketing automation platform. It’s much easier to create a new field, hide the existing one, and then move on than to clean and deal with the existing data. But when you’ve swept old fields under the rug instead of properly integrating the data, you end up with a CRM full of duplicate or triplicate legacy fields representing the same thing in different formats. Which field for “Lead Source” is the most accurate? Are sales and marketing even populating the same fields? Which field are we supposed to use for reports?

Let the chaos commence.

7 Strategies To Combat Bad Data Quality

Bring all of your data into one place

Just like assembling a complicated piece of furniture or putting together a thousand-piece jigsaw puzzle, managing your data becomes a more feasible endeavor when you have everything laid out in front of you. By bringing your data all together in one central place, you can start to see the full picture, from a bird’s eye view. From there, you can begin to evaluate the quality of various data sources, analyze where you have missing data and blank fields, see which records have been duplicated, and then formulate a data governance plan.

In the past, organizations have called on their own internal developers to custom build a central data repository for sales and marketing data, but now, new solutions like customer data platforms have entered the market to meet this need.

Define a consistent identity resolution strategy

Identity resolution (or lack thereof) is at the core of many data issues, especially as it relates to duplicate data. Duplicate data occurs when your identity resolution strategy has some holes in it, or hasn’t been applied. Defining and sticking to an identity resolution strategy will allow you to have a better handle on your data records and eliminate duplicates as much as possible.

Of course, it’ll be impossible to eliminate all instances of duplicate data with identity resolution alone, but it’ll at least keep edge cases to a minimum.

You can read about some common strategies in our blog, Identity Resolution Strategies for B2B.

Create data source groups, followed by a unified data group

Have you ever merged two records, realized you made a mistake, and then immediately regretted it? While many software applications have made merging data really simple, undoing the merge when you make a mistake becomes much harder.

To combat situations like bad data merging (and consequently losing valuable data), we often recommend centralizing your data, yet keeping your data separated using a concept we call “grouping”. Grouping lets you create holding chambers of data within a lead or account profile. Each group corresponds to a different data source. For example, you could have a group for your CRM fields, your data enrichment vendor fields, and your chat platform fields.

The next step is to create a “unified data” group that lets you surface the best data when available. For example: our unified data group populates a lead’s job title from our data enrichment vendor, Clearbit, but if a more authoritative source, like a direct form submission from the lead, were to introduce a new value for job title, we would override Clearbit’s data with the new data.

With data source grouping, the integrity of the original data sources stay intact, while allowing you to apply rules to surface your preferred data when available.

Use data processing tools

Data transformation and processing is a strategy that involves either changing or generating new data by running rules against existing data.

Standardizing data is often a good use case for data transformation. At Hull, we use our data transformation tool, the Hull Processor, to programmatically standardize data. Data transformation using the Processor is an automated process that “scans” for specific data scenarios when they appear, and then changes the data based on rules that you set. A simple example: If “GA” appears in a State field, change it to “Georgia”.

Data processing tools allow you to not only clean and standardize your data once, but also keep it maintained over time.

Implement third-party data enrichment

Data enrichment is the process of associating publicly available, third party data on a lead or customer with what you already know about them in their existing lead or customer record. Once you’ve brought your data into one place and created your single customer view, you may still have empty fields.

Data enrichment with a trusted vendor like ZoomInfo or Clearbit can augment your customer profiles with relevant information, allowing you to fill in the blanks that remain.

Whitelist certain segments to prevent bad data circulation

Bad data in one tool is a manageable offense. Bad data synchronized across all of your tools, without a proper data governance strategy, is a waking nightmare. A lot of B2B marketing teams send and synchronize all of their marketing automation contacts to their CRM, and vice versa, mainly for simplicity. But this strategy ensures that if one tool shows red flags for low quality data, the other tools will as well.

We recommend a strategy that prevents bad data from being populated across your technology stack, and the way we implement it involves using our customer data platform and creating “whitelisted” contact and account segments. Contacts and accounts in these “whitelisted” segments are synchronized to other tools, while the rest are withheld. This strategy lets us fine-tune our data flows and allow only the records with data we trust to be synchronized.

Use defensive coding as a failsafe

Hopefully, after completing the steps above, you feel confident in the quality of your data.

But, just in case, you can implement defensive coding in your marketing campaigns as a fallback mechanism. Defensive coding is a strategy that considers what would happen in a worst case scenario. For example, if your marketing email uses a variable tag for ‘First Name’, but for some reason, some of the leads in your marketing automation tool have ‘First Name’ missing, you can employ defensive coding to fill that spot with something else.

Here’s what it might look like in practice:

Hi {% if customer.first_name != blank %}{{ customer.first_name | capitalize }}{% else %}there{% endif %}!

This basically says: Plug the customer’s First Name after “Hi”, but if there’s no First Name value, insert “there” instead.

Strategies like these offer extra precaution in your campaigns, just in case there are hiccups in your data.

Get your data into gear

Poor data quality can be caused by a lot of different scenarios, but the good news is that there are concrete steps to take to begin improving your data hygiene. Got questions for us? Feel free to send us a chat anytime!

Did you learn something new?

We at Hull research how to "join the dots" between your tools, teams and data. Subscribe to follow along with more articles like this, and learn the latest trends, tactics, and techniques.

Headshot
Angela Sun

Angela brings over a decade of B2B technology marketing experience to her role as the Director of Marketing at Hull. Prior to Hull, she spent 5 years at utility data aggregator, Urjanet, where she held various roles in demand generation, marketing operations, and product marketing.