Data enrichment part IV: selecting data providers
This is part 4 of our blog series on data enrichment. Feel free to catch up if you missed the first three posts: Introducing the Data Enrichment 101 Series, Determining the Processes the Data Will Support, and Determining Your Target Market.
Now that you’re clear on your own goals and needs, next you must decide on your prefer consumption model. Here are the factors to consider:
- Single Data Vendor or Best-of-Breed
Whether you’re looking for only contact data or comprehensive data that includes firmographic, contact, social, technographic, and intent, you’ll be faced with the choice of using only a single vendor that gives you the widest coverage or assembling a collection of best-of-breed vendors. The advantages of a single vendor include:
- Lower cost
- Less integration and automation requirements
- Less confusion for users
- Fewer data standardization issues
The benefits of the best-of-breed approach are:
- Highest possible data quality, depth, and coverage
- Cross validation
- Single Source or Multi-Source
Match rate for data providers can range from under10% to over 70%. We’ve seen most match rates tend to fall between 30% to 50%. So if your preferred provider doesn’t find a match or returns a low confidence match, would you like to try another provider? Another common reason for multi-sourcing the same data is to get the necessary geography, vertical, or company size coverage.
- Cost Model
Data vendors and providers use different pricing schemes. All else being equal, the best-fit cost mode can help you maximize the ROI of your data investment. Here are the most common models and when they make the most sense:
- Credits / A-La-Carte – You buy a basket of credits and those credits can be used for different levels of services. For example, email validation usually costs the fewest credits, appending existing leads costs more, acquiring a new lead costs even more, and acquiring additional data like direct dial phone numbers and social data may cost even more. The benefit of this model is that you can maximize your budget given your desired mix of data needs. However, tracking usage and budgeting can get tricky.
- Per Record – You pay once for each record you receive. You don’t pay again if you send in the same record again. The benefit here are simplicity and that you will never pay for the same record more than once. This makes it simple when it comes to implementing the process and technology that will use this service, because you don’t have to worry about coordination.
- Per Request / API Call – You pay every time you make an inquiry, so if you send in an enrichment request for the same person three months in a row, you pay for that record three times. This makes the most sense when you are doing batch bulk enrichment. It makes the cost tracking very simple if you don’t have multiple request channels.
- All-You-Can-Eat – You rent the entire database or a subset of the database and can get as many leads and make as many inquiries as you want. This is the best choice if you want to keep your records absolutely up-to-date using automation tools.
Understanding the Data Supply Chain
It is a very very common practice within the data business to source data from other data providers. In fact, the majority of data vendors and providers do not have original data they produce themselves, but are aggregators of other people’s data. Even providers that do have original data, they often source third-party data to produce a more comprehensive offering. Here are some examples of such:
- If the data contains DUNS Number, it is guaranteed that the provider has sourced firmographic data from Dun & Bradstreet (D&B)
- Salesforce’s Data.com’s firmographic data is D&B data
- Many provider’s technographic data is from HG Data
- Social media data are often sourced from SYNTHio and FullContact
- Email validation is often sourced from BriteVerify and FreshAddress
Whether you are going with a single provider or best-of-breed approach, you should always ask the providers which third-party data have they sourced. It is import to understand what the original data sources are because:
- You should not unknowingly pay for the same data multiple times
- If you are using a multi-sourcing strategy to achieve a more comprehensive coverage, you may not get the coverage you were targeting
- This has great implication when it comes to GDPR compliance, which we will discuss in the next blog.
Each one of these sourcing relationships is likely unique. The provider can source anywhere from just one data field like the DUNS number, or the entire data set. So ask not only which third parties they source data from, but which data field are sourced from whom.