4 Steps to Cleaner Marketing Data: A Guide for MarketersAndrea Lechner-Becker / August 18, 2020 / 0 Comments
Marketers aren’t traditionally “data” people. Historically, data has been an IT function. But today, marketing technology (MarTech) empowers marketers to collect and manage marketing data on their own. This data is used to make real-time decisions and dynamically adjust messaging to align to buyer pain points. All this can happen regardless of data quality, but it only happens well when marketing data is accurate. After all, even the best campaign targeted to the wrong segment will not work. In this article, we outline the broad strokes for how to tackle dirty data challenges.
Perform a Data Assessment
You probably have a hunch you have a data problem but cleaning it up might sound overwhelming. So, start by reviewing your database. There are various degrees of detail from which you can zero in on database cleanliness.
This isn’t a step toward cleaning your marketing data yet. Instead, we want to understand the size and scope of the issues.
Duplicate records for human beings (lead, contacts, people) are the most common, and most harmful marketing data issue. As one of our consultants explains in this article, duplicates cause all sorts of problems. Some are obvious, but many stay undetected for long stretches of time, causing frustration for end customers.
Most tools include reporting and merging ability for duplicate records. Salesforce.com, for example, offers ways to create reports on duplicate records and manage them. I know it’s tempting, but try not to merge anything yet. During this first step, you’re only looking to get your arms around the data, not manipulate anything yet.
Make a list of all the fields you use to segment. Some will be compliance-focused, like country and state. Some will be messaging focused, like job role or seniority. Some will be product focused like, interest or relevant technology.
We always recommend using something like Salesforce’s RingLead Field Trip to run reports on all of your data fields. This helps you quickly figure out the percentage of records with values in any given field, and the overall health of your data. If multiple fields are populated only 1% of the time, you will naturally ask yourself if you even need the field. This should also help identify duplicated fields existing in or across objects (or tables).
Another common issue you may run into here is a mismatch for changing picklist values. A common example of this is with Country and Industry. Let’s consider Country first. We commonly see organizations that have left this field open text, which results in various values for the same single option, like U.S. vs. USA vs. United States vs. U.S.A. vs. you get the point. Likewise, depending on your historic data governance around Industry, we typically see disparate values which originated from list imports, individual laziness (selecting the first option on the picklist), form submissions or changing methodologies around granularity (SIC 2, 3, 4 or not at all).
Again, you aren’t solving for any discrepancies yet. You’re only looking to collect data and understand where the issues reside.
Step 1. Categorize, Standardize & Prioritize
Once you’ve identified where your issues lie, it’s time to get started on solutions. For step 1, you will aim to categorize your data for the purpose of making standardization decisions. Then, you’ll prioritize the actual clean-up activities.
For the first piece around categorization, you’ll likely find these broad categories in your data:
- Question asking. Sit down with the people who actually use (or attempt to use) your marketing data on a daily basis. Ask people from each department, the following questions and look for universal patterns to emerge.
- In x-system what data do you use?
- What data do you think is a total waste?
- What data do you wish you had?
- Decision-making. Often, a standard for a field does not yet exist. Returning to Industry, sales and marketing both use the field for targeting and segmenting accounts. Everyone needs to be on the same page about what values, and what level of detail, is important. We typically see organizations either use SIC codes or a vendor to make this decision. The vendor decision is an easier lift, as it would basically be saying something like, “ZoomInfo’s industry list looks like this and they update our data automatically, so let’s just use their list.” In return for that ease, there is less room to customize the options for your business.
- Systematic inputs. If you have forms submitting inaccurate data (e.g. U.S.A. instead of US), you need to map all those input sources and what they’re currently doing. Same for any integration points you have. If your product collects information, for example, and pushes it to another system, it may have different values or unique fields. Document these explicitly.
- Human inputs. The companion to systems, if you have humans creating the data, ensure you’ve identified what options are allowed in their systems, but also what processes they have. If, for example, you require industry for account creation, but someone calls in and you don’t have flows for your sales reps to ask for company during that first call, it will hamper your ability to maintain the consistency you’re looking to achieve.
Once you have the granular data issues documented and standards agreed to, it’s time to prioritize and plan for when and how they’ll get corrected.
Step 2. Remove Duplicated Data
You’ve done the talking, the deciding and the prioritizing, so now’s the time to get to work.
BUT FIRST: A nugget of wisdom. Before you delete anything, always export your entire database (every object, every field, every record) first.
Then, remove duplicates in all the areas you found, from field values and fields to whole records of leads and accounts. The gold standard for duplicate merging in Salesforce is still DemandTools.
Step 3. Capture & Hygiene
You identified the systematic inputs in step 1.3. above. So, follow your prioritization and start cleaning up these systematic inputs. From marketing forms, to product profile inputs, start updating the systems to ensure all the hard work you’ve done to combine your duplicates doesn’t just go back to being dirty.
To this end, maintain a consistent approach to hygiene. Maintaining a clean marketing database is far from a one-time project, but instead requires diligence. As someone who routinely skips my weekly data cleanliness reporting, I recommend saving, at least monthly, half-day blocks on your calendar for clean-up. It’ll lead to happier customers and ultimately save you some money on database size.
Step 4. Append
The rule of thumb is that the less fields on your form, the more conversions you’ll get. So, although you might be tempted to ask for a lot, leverage technology to help you maintain cleanliness and also acquire more leads. Look at ways to automatically append data, like DiscoverOrg’s nightly syncs (which is what we use), so you’re asking the bare minimum from your customer.
Bonus points for the appending of additional records as well. Ideally, you’ll want to add complimentary contacts that will enable your sale. Likely, that means adding C-level contacts as well as more hands-on users and influencers. The sooner you start nurturing the organization as a whole, the sooner you land the account. If you’re curious how to identify your buying committee, take a look at this article: “Buying Committee”
Yes, it’s a lot of work. Yes, it takes time. However, there is no denying the fact that a healthy marketing database will build confidence with both internal teams and buyers by ensuring that your message reaches the right people and in the manner they want to receive it. Clean marketing data results in better buyer relationships and revenue. Worth it? Heck yes.