Clean up your customer database: detect insolvent and defunct companies automatically
Before you reactivate dormant customers, clean up your customer database. How an agent detects defunct companies and puts unclear cases in front of a person.
Your sales back office starts a reactivation: customers who have not ordered for some time are to be contacted again. The first e-mails bounce as undeliverable. On one call, a completely different company answers the number. A third customer ceased trading months ago – it just was not recorded anywhere in the system. Before you reactivate dormant customers, it pays to clean up your customer database.
Why you should clean up your customer database before reactivating
A dormant customer is not automatically a lost customer. Some business relationships are only asleep because nobody got in touch. That is exactly where customer reactivation comes in. But some of the dormant records describe companies that no longer exist.
A look at the figures shows that this is no side issue: according to the German Federal Statistical Office, there were 24,064 business insolvencies in Germany in 2025, 10.3 per cent more than in the previous year and the highest level since 2014 (Destatis, 13.03.2026). Insolvencies have many causes. For your customer database, they mean one thing above all: records that were correct yesterday can lead nowhere today.
If you keep contacting such companies, you lose twice. Your staff spend time on calls that cannot achieve anything. And your metrics paint a distorted picture of how well the reactivation is working, because they include customers who can no longer respond at all.
Two sources that reveal a lot
Whether a company is still trading can often be read from publicly available information. In the REVAGO system, an agent checks two sources for this:
- Commercial register data: entries on insolvency, liquidation or deletion are a clear signal.
- The customer’s website: the home page and legal notice show whether a company is still present – or whether a closure of the business has been announced there.
The two sources complement each other. The register shows the formal status, the website the actual presence. A company can still be registered and yet have ceased trading. Only together do both give a reliable picture.
If neither source gives a clear picture, the agent can fall back on a web search. A fixed budget limit applies to this, so that the check stays economical. Why such limits are part of running AI agents is explained in Reducing and controlling AI costs.
What the agent decides and what a person decides
The check alone is only half the work. What matters is what follows from it – and who decides.
If an insolvency, liquidation, deletion or closure of the business is confirmed, the agent marks the customer and their contacts as “not relevant”. Ongoing outreach is stopped. Nobody has to follow up these cases one by one by hand.
If the situation is unclear, the agent does not decide. It puts the case in front of a person. The same applies to companies abroad. That is human-in-the-loop where it is needed: the agent handles what is unambiguous, and what is doubtful stays with your staff.
This division of labour follows a principle that runs through the REVAGO system: agents act within approved scopes. What an agent carries out itself and what it only prepares is defined by you in the mandate. More on this in Agents: AI agents as the new workforce in sales.
Human decisions stay in place
A cleaned-up customer database only helps if it stays clean. If a member of staff has reviewed and classified a customer, the next data sync with the ERP system must not quietly reverse that decision. In the REVAGO system, human decisions survive the import.
This is part of a bigger question: how reliable is the data your agents act on? Duplicates, quotes instead of orders and bounces are covered in CRM data quality.
Clean-up as part of reactivation
Detecting defunct companies is not an end in itself. It makes customer reactivation more effective, because outreach and attention go where a business relationship can be revived. What the whole chain looks like, from the first contact to the follow-up call, is shown in B2B customer reactivation.
That is why we design this check starting from the result. In AI-first coaching, you find out whether and how AI agents can secure and grow your revenue – and which customer groups are in focus. During implementation, we define with your functional leads which cases the agent removes from outreach itself and which a person reviews. In operation, the check keeps running, because companies do not close according to plan.
Your next step
How many of the dormant customers in your system still exist? If you cannot answer that question today, it is a good starting point for a conversation. Before you start a reactivation, check your REVAGO potential.
