We use a lot of data in marketing, and it can supercharge all that you do when used appropriately.
There seems to be some inconsistency to how we talk and think about it. There is 3rd party data that you can get from various vendors. I highly recommend testing out data from a number of vendors (for quality as well as usefulness). Like at a used car lot the salesperson will have you believe there isn't a lemon among the bunch. I've had a vendor gaslight me... she tried to tell me the experience I had with a certain source of data was not the experience I had. Keep in mind they may believe all that they tell you, but they are salespeople not practitioners and don't have hands-on experience setting up, using, and troubleshooting the data.
Let's categorize some of the most popular types of data you can source from 3rd parties, so we have a framework to approach data.
Firmographic
Definition: Most B2B marketers are familiar with this type of data. These are characteristics that describe the company / organization (Revenue, Number of Employees, Industry, etc). These normally don't change much over time.
Use: This data can help you validate the characteristics of companies you sell well into to and can thus find more that are similar for your various segmentation of the Total Addressable Market (TAM) and company scoring needs
Considerations: A complexity is how do you define an organization, and it's relationship with a parent... that will determine at what level you need the data
For instance Sony is a multi-national conglomerate, at the simplest level you may sell at the region or country level and want to have data (i.e. revenue) aggregate at that level
Depending on how your sales force sells, you might want to aggregate to the Brand / Group level... Sony for instance is well known for consumer electronics, but they also have very large businesses in professional electronics, music publisher and record label, film studio, Game publisher, etc.
The grouping of the company is important to know which brands / groups you will do well to focus on and which team should work the deal (specialized sales team based on industry etc)
Be sure to find out which vendors can provide which data at the level you need (it's not uniform across data types even with the same vendor)... often at a Global parent level is not enough and gives you frustrating false signals or noise rendering the system untrustworthy by users
Technographic
Definition: Data on technologies used at prospect companies / organizations, often grouped by categories and subcategories. Depth and coverage depends on the vendor. They change in the yearly timescale and in only accelerating with more SaaS solutions
Use: This can help you find prospects that are already using competitor solutions or complimentary solutions (if those are known) and help you find them if it's not known. Again great for initial TAM segmentation and company scoring
Considerations: Often this type of data is aggregated at the global levels (web domain) and sometimes at a brand level but rarely at a country level for instance (be aware of the limitations as you use to segment)
In front of firewall tech is easier to get accurate data about as clues can be scraped from prospect/provider websites vs behind the firewall technology
Because data is collected from scraping and inferring the existence (i.e. scanning job postings and the requirements etc), the accuracy can be suspect and there is rarely an indication of how recent that tech was known to be used
3rd party activity
Definition: This is often called Intent data, but really it's just a collection of 3rd party activity grouped by topics/Interest. The categories range by vendor and depth of details and breadth of collection is highly variable and unclear. Some have data about downloads of e-books and have very detailed info about how many users and when but they have very little breadth as it is only download form their site. While others collect data based on web visits of pages they have tracking scripts on (i.e. Ad Serving) and pull keywords from the pages to assign a topic. These usually only give you aggregated data in the form of a processed score. While this is broader in breadth in terms of audience surveyed, the accuracy of the keyword approach come into question. To try to address some vendors will use Natural Language Processing (NLP) techniques to extract what the page is about.
Use: This can be a good tool help you increase your awareness of activity in the market and identify companies / organizations that are in market but have not reached your web properties and assets yet. Another interesting use case is detecting research at customers (perhaps a sign they are looking at competing solutions)
Considerations: You will want to back test know outcomes (deals that made it into funnel, those that were in touched by sales but had no need / interest) and the 3rd party activity have show up ahead of the deal and arriving on your assets and properties and that is its greatest value.
Vendors claim relationships with thousands of web sites, but you have to wonder what are those sites and is my target prospect on those sites looking for this kind of info there
Depending on the provider, you may only get a score and a date for that score and getting historical info or more detailed info may not be available, this makes back testing and modelling impossible to prove value until after a history of use where you store the historical snapshots on your systems... Activity data is time sensitive.
Demographic
Definition: This is Contact Data another area most are familiar with. The data is about the person primarily from a profile perspective. Though some platforms let you use person level 3rd party activity for ad targeting but not for data enrichment.
Use: Generally this is most useful for getting links to the right companies and getting person's roles and contact info. Another related use case is identifying contacts that you are not aware of but fit persona's you care about
Considerations: This is also a mixed bag by vendor and the data's recency and accuracy at each field level... some have great info for email but not phone etc... also coverage of the accounts your prospects are at may vary greatly. Best to test a random sample.
For contact discovery, you can provide titles and roles you are trying to source or the vendor may have predefined personas you can leverage. In some cases you can request customer personas to be created.
Testing your data sources
To evaluate vendors, you will want to test each type of data you are interested in. Here is a starting point.
Account Data Append
Create a random sample of you account... you may want to create sample with relatively equal numbers buy region (i.e. Americas, EMEA, APAC)
Even if you don't have business spread evenly across regions... this will let you evaluate data in each region
Within each region you might want to create equal samples representing company segments you sell to (i.e. by revenue band)
This will let you evaluate the data quality by company size etc as it is far easier to get at data for the largest companies but not as easy by brand or smaller companies as there is less publicly available info
Contact Discovery
Again using a randomized sliced list of companies, you can ask for getting samples and counts of contacts that match your criteria (titles, personas, etc)
The counts they may be able to provide across your TAM giving you a sense of coverage across the slices of you TAM (i.e. region / company revenue)
Contact Data Append
Similar to the Account data append, provide the vendor with a list of contacts you need enrichment for
Evaluating Data Quality
Evaluate each slice of sample records for:
Coverage of data (type/fields)
This is mostly to see what data is populated and types of fields available in each of the categories (Firmographic, Technographic, 3rd Party Activity, and Demographic)
Accuracy/Recency
You can review each record or a sample of companies you team know well
Sometime data evaluation will be hard with a large number of records. In that cause you might want to identify records where most vendor sources agreed and focus on the one that didn't look like the others to see if the data is newer/better or if they are the of man out with old in accurate data. This might help reduce the records needing manual follow-up.
Coverage Contact Discovery
Contact coverage (by roles / personas we care about)
the samples of contacts they return to you will give you a sense of the quality of the contacts they are able to identify... if it's agreeable have you sales team or telemarketing call and validate info and fit. You should also see how many were already known to you... as the value is finding new prospects.
Accessibility
How you are able to access the data greatly affects how you can leverage it (API / embedded app / platform interface)
some platforms enrich you DB while others you can only use the rich data (i.e. Technographics / intent) on their platform to generate list of matching accounts/contacts
the reason this is important is it determines if you can use that same data easily for additional processes (reporting, model building, etc)
You will need to weight which are most important to you
Cost / benefit
Picking the lowest cost solution is problematic as you usually give us data quality... but picking the highest quality can be too expensive to use at scale
This is where you might want to weight the cost and quality and decide to use the low-cost good quality solution for the bulk of your data enrichment but a high-cost solution for a smaller subset of records giving you more flexibility and quality where it is most needed
Obscure problem child data
include a large number of problem records to stress test their data, and again have a team validate the info
These can come from routing issues or records that were dropped into round-robin because an account match wasn't possible... or areas where data is sparse like public sector
This can be the basis of your scorecard, and you will need to put weight on each element and criteria for the level of score in each.
First Party data
Data from 3rd parties is an important part of our data mix, but we shouldn't forget about our 1st party data as we have the most control over its depth and breadth. The following isn't an exhaustive list but ones I've found to be useful over the years.
Individual Activity
Definition: This includes web activity on properties you own as well as responses to campaigns and attendance at events (live and digital)
Use: Activity can be used as a proxy for interest, trigger for follow-up, or for modelling (persona predictions, Interest scores etc). These activities inform models and the sales team as to who to follow-up with and how to position the door opener script.
Considerations: Beyond the immediate use when trying to understand the contact, aggregating the activity info at the Account level can help create a picture of what is happening at the account as a whole and the overall interest.
If you are using a Buyer Group funnel, grouping the activity and scores at this level can be used as an additional rule/gate to progress the Buyer Group for follow-up
Deal Details
Definition: Includes the usual, products / services in consideration and the deal amount. But also includes details like stage and time in stage as well as people involved (internal and prospect)
Use: This is very use in triggering other activities and notifications or suppression of marketing activities for deals deep in negotiation. When look back at a quarter or year this is really useful in determining sourcing credit and influence credit (see post How to translate business goals to effective Marketing targets)
Considerations: One area of data that is consistently under populated are the people involved on a deal at the prospect organization. There are tools that scour emails and meetings of your sales team to populate people involved in deals. Automation help reduce the manual work, but you give up some accuracy. Add a layer on manual validation will help with accuracy.
Sales Activities
Definition: These are the records of the emails, calls, messages, meetings, etc that are part of vetting and moving a deal along
Use: Knowing how and what content was sent can be used to track if a deal is active, getting enough attention, and is progressing as expected. This can prompt or prevent marketing communications/promotions as desired and agreed upon. It can also help you identify best practices when paired with deal stage info, to know what works well at which stage
Considerations: Again automation is key here... if tools can automate the logging of this rich data... sales people can focus on building relationships and progressing deals rather than administration of the deal process
Products Owned / Entitlements
Definition: This is a representation of what the customer has... product / services etc and can include details about contract start and end dates etc.
Use: Having start and end dates help you trigger customer onboarding programs and renewals processes or checkpoints to ensure the customer sis getting value from your products and solutions. Know what they currently have and how much on it can also help you model / understand if they are likely to buy more or buy a complimentary product/solution. Tactically you can adjust you content and targeting based on know what they are already customers of.
Considerations: This should tie into your provisioning systems, customer communications, and customer success systems. You want to ensure the data is sourced appropriately rather than having conflicting sources of data that can get out of sync.
Support Cases, Surveys, and Success
Definition: I know this is a little unfair to group all these important data point into one category but... This not only includes the support requests made by customers but should also include any feedback from surveys or interactions with customer success.
Use: This is are barometer into the customer and when used with all the other data we have we can model for attrition and upsell opportunities. This can also help us find customers that are big fans to power and advocate and reference program.
Considerations: There are entire platforms dedicated to customer success. We should tie into those systems for this data to ensure we aren't create unlinked copies of data that will become stale and out of sync.
Data is a raw material and what we do with it to create best in class customer experiences is only limited by our imaginations. What types are essential to your processes? What types of data are you starting to look at or aspire to use? Join our community and share in the comment below.
