Research an account without adding more people

Fresh research should not grow your contacts

You may need a fresh view of a company without another person in your contact list.

Maybe you already know who decides. Maybe the company changed its site, and you want fresh evidence before a call.

On October 23, 2026, we added a clear choice to Mach's company research. Research can now run without adding new contacts.

The other choice still does both. This keeps research to the job you asked for. It does not mean more contacts are always bad.

A refreshed company and a longer contact list are different results. Showing the difference lets your reps choose the work, not find the side effects later.

This is an account of our own October 23 and 24 changes. It is not a study of customer results.

Choose the job before starting it

The research menu now has Research and Research and add contacts. The first studies the company and who decides there, without adding contact records.

The second also adds the people it finds, with their titles and public profiles where available.

Learning who decides is not the same as adding them as contacts. A company brief can describe a role without adding a record.

Research paths that do not offer the new choice keep their old behavior. Older ways of working still add contacts.

Check the exact path you use. Do not assume the new choice changes every run. Our guide to the people who decide explains why a title alone does not prove authority. [1]

A made-up refresh with two paths

Imagine Tarang Labs, a made-up software team in Pune, already talking with Mara Ellis at Alder Creek Bank. Jamie Carter owns the deal and knows who joins this week's call.

The bank announces a new loan system for March 2027. Jamie wants fresh company facts and better questions, without adding new people.

At another bank, the team has company evidence but no one to write to. There, research plus new contacts may be the better job.

The table compares these made-up cases. It does not report customer behavior, added contacts or measured savings.

Both paths still need checking. A newly found person may have an old title. A news item may describe a plan, not finished work. The research choice does not remove those doubts.

Write down the contact count before you start. Afterward, read both the company brief and the people list. If Jamie picked research only, new company facts are not new contacts. Judge the brief on its own evidence.

A retry must remember what you chose

A choice on the first screen is not enough if later work forgets it. Work in the background can resume after the first request has ended.

The October change stores the contact choice with the research run. Supported retries read that stored choice instead of guessing again.

The step that adds contacts then checks whether this run may add them. Research-only runs skip it.

Tests check the chosen action, the saved choice and whether extra people are added. They do not measure how often it runs or how well.

When you judge a similar option, ask what happens after a retry. A choice that lasts only until the first failure is not a real limit.

Check the result as well as the menu label. Research only should not start adding contacts when stopped work resumes. Test that before you rely on it. The saved choice and the contact list should agree.

Separate research cost from follow-up work

Here is a made-up cost example. Checking 10 contacts you did not need, at 3 minutes each, takes 30 minutes.

At $60 an hour, that is $30 of labor. The example does not estimate what Mach charges for research or contact data.

If you need those people, the check may be useful work. If you only wanted company facts, the wrong job created it.

Track both kinds of effort in a trial. Record time spent checking the company brief apart from time spent checking new people.

Do not turn the example into a promise of savings. Your team may already have a clean contact list, or a good reason to grow it.

For a small internal test, count contacts before and after a research-only run. Also read the brief itself. An unchanged count does not prove the research was useful. Those are two checks for two results.

The mistake was bundling two useful jobs

What we got wrong was making new contacts an automatic part of company research. Each job can be useful without the other.

The choice is clear where the new research actions appear. Other research paths still add contacts as before.

This change does not claim every run finds a new person. It only controls whether this run adds them.

Before your next refresh, write the result you need in one sentence. Do you need fresher company facts, more people to write to, or both?

Pick the matching job. Then check the evidence before you treat any new fact or person as ready.