Preparing for a data evaluation in under an hour

The evaluation is the step where an estimate becomes a real number. It is read-only, it takes under an hour of your time, and that hour goes faster if you have done a little homework first. Here is what to gather, what to decide and what to expect.

Before the evaluation, any figure is a range based on what you have described: headcount, years in operation, the tools you used. After it, there is a count of what actually exists. That means how many years of history each system holds, how many people appear, how many tools overlap and how often the same work shows up in several of them.

That count is what an offer is based on, and it is why nobody gives a firm number before it happens. Two companies with the same headcount and the same tools can hold very different datasets, depending on how far back their history goes and how consistently their teams linked their work together.

It is also where you learn what is not there. A tool you thought held five years of history may only keep two, or a migration may have left a gap. Finding that out early keeps the offer grounded and avoids surprises later.

The evaluation is read-only. Nothing is changed, deleted or moved out of your systems at this stage, and no data is delivered to anyone.

Most of the friction in an evaluation comes from hunting for access in the middle of it. A little preparation usually removes it. Before the session, find:

• A list of the systems in scope: chat, email, docs or wiki, ticketing and project tools, code hosting, CRM, support desk and any finance tools
• The admin or owner login for each system, or the name of whoever holds it and a time they are available
• Rough dates for when each tool was adopted, and any migrations, such as a move from one tracker to another

The third item helps more than people expect. If you switched project tools four years ago, knowing that up front tells the evaluation where to look for older history. A rough headcount history helps too: how many people worked at the company at its peak, and roughly how many over its whole life. Headcount and years of history are two of the main inputs to value, alongside the number of systems and how well they connect.

If a former employee held the admin role in a tool, recover or transfer that access now rather than on the day. For companies that are winding down, this is worth doing urgently in any case, as explained in before you cancel Slack.

Scope and exclusions are set by you and annexed to the contract before any data moves. It is easier to make those decisions before the evaluation than during it, because they affect what gets counted.

The defaults already cover a lot. HR, payroll, medical and legally privileged folders are excluded automatically, and direct messages are left out unless you choose to include them and have the right to. Beyond that, think about specific clients with strict confidentiality terms, channels or folders that touched a sensitive deal, and any repositories that contain code you licensed from someone else.

Writing these down as a short list makes the conversation quick. Our post on exclusions has examples, and customer confidentiality clauses and your data covers the contractual side. If you are unsure about a particular item, leaning toward excluding it is a reasonable default. It keeps the decision conservative, and its effect on value can be discussed openly before anything is signed.

The session itself is mostly about granting read-only access, system by system, and answering a few questions about how the company worked. Expect questions such as which channels were most active, whether teams referenced ticket keys in their commits, and whether any tool held data from a separate business line.

You do not need to clean anything up. Old channels, half-finished docs and abandoned projects are part of how real work looks, and that is part of the value. Credentials or card numbers that people pasted into chat over the years will be removed during anonymization, so there is no need to search for them yourself. We explain how that works in secrets in the archive.

It also helps to have one person on the call who knows the company's history. That might be a founder, an early engineer or an operations lead. They do not need technical depth, just a sense of which tools held the real work in which years.

Once the count is complete, you receive an offer that reflects what is actually there. Typical licences range from $20K to $400K+. If you decide not to go ahead, nothing has left your systems, and you can revoke the read-only access. If you accept, anonymization happens in-house: every person, company, address, email, phone number and account becomes a consistent token across every tool, and no third party touches the raw data.

You then review an anonymized sample before anything is delivered. Payment goes to the company entity in two to four weeks, and the raw archive is deleted within 30 days of sign-off as a contract obligation, with a deletion certificate. Across the whole process, the time asked of you is about an hour. For the full timeline, see how long does a data licence take.

If you have your list of systems and know who holds the admin logins, you are ready. Book a 15-minute intro call and we will schedule the evaluation from there.