Research columns: from a list of names to a list of reasons

A prospect list tells you who exists. A researched list tells you who to contact first and what to say. Here is how to design research questions that actually change what you do next.

GTM Radar team3 min read

Every prospect list starts the same way: names, companies, job titles, maybe a website. It might come from a CSV someone exported, from a connected source, or from a list you built by hand. Whatever its origin, it answers only one question: who could we contact?

It does not answer the questions that decide whether a message will land. Is this company changing in a way that makes the timing right? Is this person the one who feels the problem? What would be worth saying in the first line?

Those answers usually live in a rep’s head, or in a dozen open browser tabs, or nowhere at all. Research columns put them on the list itself.

One question, asked of every row

A research column is a question you would ask about a single account, written once and answered for every contact on the list. Instead of researching account by account, you decide what you want to know and the research fills in the answer row by row.

The result is a list you can sort by the thing that actually matters. Not “companies in our industry”, but “companies in our industry that are hiring for the team we help”.

Write questions that change a decision

The most common mistake is asking questions that are interesting but do not change anything. Before adding a column, ask: if I knew the answer, would I do something different?

Questions that tend to pass that test:

  • Timing questions. Is the company hiring for a relevant team? Has it announced a new location, product or market?
  • Fit questions. Does the company sell to businesses or to consumers? Does it appear to run paid advertising today?
  • Role questions. Does this person own the area we help with, or are they likely to pass it on?
  • Opening questions. What is one recent, public thing about this company that a first message could mention?

Questions that usually fail it: broad summaries (“tell me about this company”), anything you would not act on, and anything that needs private information you should not have.

Keep answers short and checkable

A good research answer is short enough to read in a table cell and specific enough to check. “Yes. Two sales roles posted this month” is useful. A paragraph of general background is not, because nobody will read it across two hundred rows.

When you write the question, say what form you want the answer in: yes or no with a reason, one sentence, a single name. Consistent answers are what make a column sortable.

And treat answers as research, not fact. Before an answer goes into a message, the person sending it should be comfortable standing behind it.

From reasons to next steps

Once a list carries reasons, the next step is obvious for most rows:

  • Contacts with a strong timing signal go into a follow-up sequence, where each message is prepared and waits for your approval.
  • Contacts with a call already booked get a meeting brief on the attendees and their company.
  • Contacts that do not fit yet stay on the list, researched and ready for when something changes.

The list stops being a pile of names and becomes the start of the work.

How this works in GTM Radar

In GTM Radar you can build a prospect list in the workspace, upload a CSV, or pull contacts in from connected sources, then add AI research columns to enrich every contact. Your lists sit in the same workspace as follow-up sequences, meeting briefs and ad audiences, so the research you did once is there when you need it.

Nothing is sent from a list. Researched contacts move into follow-up only when you add them, and every message still waits for a person to approve it.

Start with three columns

If you are new to researching lists this way, start small. Pick three questions: one about timing, one about fit and one that would give you an opening line. Run them on a list you already know well, and check whether the answers match what you would have found yourself.

When they do, you will have turned an afternoon of tab-switching into a sorted list with a reason next to every name.

Run your pipeline with a person at every gate.

Talk to us