By Touchbase Technologies · Published 2026-11-05

What to check before saving a prospect to your customer list

A business can match your industry filter and still be a poor prospect. Before saving it as someone to contact, check what it does, which part of your product could help, where that conclusion came from, and whether you have an appropriate way to reach a person there. Keep the evidence separate from your interpretation. A plausible problem is a reason to ask a question, not proof that someone wants to buy.

This worksheet is for a founder reviewing a first customer list. It produces three outcomes: ready for a relevant introduction, needs more research, or excluded. The example uses a fictional equipment-rental workflow product; the companies, counts and decisions are illustrative, not Touchbase customer results.

An original prospect review diagram with four evidence checks: business fit, observed workflow, appropriate contact and send eligibility.

Write the product boundary first

Start with one sentence about what your product actually does. For our example: “Our software lets a rental desk collect equipment-return photos and flag items that need inspection.” That is specific enough to test. “We help businesses grow with AI” is not.

Write the exclusions next. This product does not calculate insurance claims, replace an equipment-maintenance system, or prove that a renter caused damage. Those boundaries prevent a researcher from selecting any business with equipment and inventing a need. If an account only rents digital assets, it is outside this example even though its website uses the word rental.

Your audience definition should describe a workable situation rather than a broad identity: physical equipment, a return process, and someone responsible for the desk or operations. Geography is another boundary. Touchbase's business search defaults to the United States; use an optional city or state when your offer genuinely has that restriction.

Keep four fields distinct

Field	What belongs there	What does not belong there
Observed fact	The business publishes a page describing weekend equipment returns	A guess that staff often lose equipment
Source and date	The specific page and when you reviewed it	An unattributed AI summary
Fit hypothesis	A return-photo workflow might support that process	A claim that the company needs your software
Next question	How are return photos attached to a rental record today?	An assertion that the current system is broken

An individual can also be a customer. For a creator tool, the relevant evidence might be a person's public portfolio and the kind of work they publish. Do not force every offer into a company-size filter. At the same time, public professional information does not establish permission for every marketing use or justify collecting private personal details.

Review a fictional three-account shortlist

Account A describes equipment rentals and a staffed return counter. Its website names an operations contact. Save the relevant page and record the role. The account is suitable for further qualification, although nothing on the page proves that its return process is difficult.

Account B is a directory containing dozens of rental listings. It appears in a search result because it repeats the right terms. Exclude it from this campaign: the directory does not operate the rental workflow your product serves. Record “directory, not operator” so another researcher does not add the same account tomorrow.

Account C sells equipment but does not describe rentals. Mark it for research rather than silently changing your audience definition. One useful next check is whether its own website has a rental service page. If there is no supporting evidence, leave it out. A smaller list with documented fit is more actionable than a large list that requires the sender to invent an opening line.

Separate contact discovery from email verification

Finding a name and finding an address are different tasks. Check that the role fits your question and that the public source actually links the person to the account. Do not generate a likely email pattern and treat the result as a confirmed mailbox.

Hunter's verification documentation distinguishes valid, invalid, accept-all and unknown results. Accept-all does not confirm that the specific mailbox exists; unknown can mean a server did not answer a verification attempt. Even a valid result cannot guarantee delivery later. Retain the status and the time checked instead of reducing every result to a green check mark.

For a small initial campaign, a practical review policy is to exclude invalid addresses, hold uncertain ones for review, and verify eligible addresses close to the intended send. That is a proposed operating rule, not a promise about how every verifier or Touchbase workspace is configured. An unsubscribe or previous refusal must remain a separate reason to exclude a recipient, even when their mailbox is valid.

Make the saved record useful to the sender

A useful record needs enough context for someone else to explain the choice. Include the account or person, product being marketed, observed activity, source, review date, fit hypothesis, relevant role, contact source, verification status, owner and next action. Leave an uncertain field blank with an explanation; do not manufacture certainty to complete a form.

For Account A, the next action might be “Ask the operations contact how return photos are linked to contracts.” The draft could open: “Your equipment-rental page describes a return counter. I'm working on a way to attach return photos to each rental. Is that something your desk already has covered?” Only use that opening when the page actually supports it. Avoid unsupported claims about losses, delays or dissatisfied customers.

Review the list before judging the message

Suppose you inspect 20 candidate records. In this fictional review, eight have a clear workflow match, five are directories, four have no rental evidence and three have insufficient contact information. The useful finding is not a conversion forecast. It is that only eight records are ready for the next qualification step, while the others need exclusion or research.

Keep those reasons. If most rejected accounts are directories, improve the discovery criteria. If suitable businesses consistently lack a relevant contact, work on contact research before writing more copy. If recipients say the workflow is already solved, revise the product hypothesis rather than assuming the subject line caused the problem.

Use Touchbase for the research-to-draft handoff

In Touchbase, add your website and review the product description and intended audience before searching. A generic product label will give the researcher a weak starting point. Review the results and the reasons for fit; a match is a potential customer, not evidence of purchasing intent. Prepare a first email using facts you can support and review it before sending.

For the next stage, use the research-to-CRM handoff guide to decide which fields your CRM needs. Check the current plans and allowances before planning a campaign; sending also depends on a configured mailbox and provider limits. Free usage is limited, with paid plans or credits available for eligible additional usage.

A final approval question

Before moving a record into a send-ready audience, ask: “Could another person open the source and understand why we selected this prospect without relying on our assumptions?” If the answer is no, the next job is research. A good list makes the first conversation more relevant; it does not promise that the conversation will happen.

Prepared with AI assistance and checked against the sources linked in this guide.