Most explanations of data enrichment list the fields you can add. Fields are the means. Below: which data points exist, then the decision they serve, and how an AI column turns dozens of fields into a keep or drop verdict for each row.
What is data enrichment?
Data enrichment starts with a record you already have, often just a company name or domain, and adds what you do not know: how big the company is, what it does, how its teams are sized, what it runs and what it is hiring for. The added fields come from business registers, company websites, job portals and data providers.
Enrichment also repairs records. Poor data is expensive: Gartner estimated in 2020 that poor data quality costs organisations an average of $12.9 million a year, and in Experian's 2019 research companies estimated that around a third of their data was inaccurate. G2 lists data cleaning and enrichment as a standard feature of sales intelligence tools.
Which data points can be added to a company record?
Dozens. They fall into a few groups, from slow-changing basics to fast-changing signals.
- Basics: legal name, domain, location, industry code.
- Size: headcount band, revenue band, individual department sizes.
- Self-description: what the company says about itself on its website.
- Technologies: the tools it runs, read from the website and from job ads.
- Hiring: open positions, and positions advertised in the past.
- Contact data: phone numbers and emails, usually handled as a separate step called contact enrichment.
The basics describe every company. Department sizes, self-description and hiring are the fields that usually decide whether this particular company needs what you sell.
What decision does enrichment serve?
One decision: is your product relevant to this company or not? Leadspicker's knowledge base puts it in those words: the point of all the data points is a single decision about relevance.
Seen that way, enrichment is not about collecting as much as possible. It is about collecting the few fields that answer the question, for every row, so the list can be cut. A list of companies becomes a list of relevant companies. Fields that never change the answer are cost without value. The same logic decides what to enrich first: company fields before contact fields, so you never pay to find a person at a company you will drop.
Leadspicker illustrates the effect as the difference between a team processing 500 leads and a team processing the 60 leads worth processing. The numbers are an illustration, not a measurement; the cut depends on your criteria.
How does an AI column turn fields into a verdict?
An AI column reads the enriched fields of each row and applies a written rule to them. You choose which columns it reads, write the criteria as a short prompt, and it returns a verdict per row.
- Map the inputs: for example department sizes, website description and open job ads.
- Write the criteria: "Keep if the company sells B2B and has a sales team of at least five. Keep if it is hiring for sales or sales development now. Otherwise drop, and give a one-sentence reason."
- Run it on every row: the output is keep or drop, with the reason next to it.
- Check a sample: read a handful of verdicts by hand, tighten the prompt where it was wrong, and run again.
The reason column matters as much as the verdict. It lets you audit the decision and reuse the reason as the first line of a message.
The cost of poor data is documented. Gartner estimated an average of $12.9 million a year per organisation (2020). Experian found companies believe about a third of their data is inaccurate (2019). G2 treats data cleaning and enrichment as a standard feature of the sales intelligence category.
How often should records be re-enriched?
Re-enrich before each campaign, and match the rhythm to how fast each field changes. Industry and location rarely change. Headcount moves over months. Job ads and technologies can change in weeks. A record enriched once and reused for a year is mostly accurate on the slow fields and wrong on the fast ones that matter most.
A simple rule: refresh signal fields such as job ads right before you contact a company, and refresh the basics when a record has not been touched for a long time or a message bounces.
Leadspicker enrichment adds dozens of data points per company, including company size, department sizes, website self-description, technologies and open and past job positions. An AI Intelligence column then takes mapped inputs and a classification prompt and decides what is relevant.
Frequently asked questions
What is the difference between data enrichment and lead enrichment?
Data enrichment adds facts about the company. Lead enrichment is the wider job of turning a lead into a usable profile, which includes company data, contact data and a fit decision.
Which fields matter most?
The ones that change your decision about relevance. For most B2B products those are department sizes, the company's own description and its hiring, more than industry codes or revenue.
How often should data be re-enriched?
Before each campaign, at least for fast-changing fields such as job ads and technologies. Slow fields such as industry and location can be refreshed less often.
