Home/Learn/How to Choose a Growth Partner

Find growth help that brings in buyers

How to choose a growth partner starts with asking for evidence you can inspect before signing. Request past work separated into completed changes, findings and plans, plus a failed test and the decision that followed it. Those two requests help you judge the work before anyone can show results for your company.

Updated November 2026

The proposals are on your desk: an agency, a consultant and a freelancer each offer to help your company grow. You could also hire your first marketer. You have clients and plenty of product and sales work, but no marketing lead to help choose. So you are comparing promises before any of these people has done work for you.

Google Search Central tells people hiring an SEO to ask about previous work and success stories. You may already have those stories attached to the proposals. Use the conversation to examine what they leave unclear: which changes were finished, which problems were only found, and what happened when buyers did not respond.

Choose a growth partner on checkable work

Ask a candidate to open a page it finished and show a problem it found. You can inspect both before buyers respond, as long as the record distinguishes what each establishes.

Use four kinds when reading that record: measured results, completed changes, findings and planned work. Results report what a measurement showed. Changes record work finished, and findings describe problems or opportunities discovered. Planned work says what someone intends to try. Keeping these kinds separate helps you see what each claim can support.

For a completed change, ask to see the thing that changed. A live page gives you copy to read; a repaired form gives you a repair to inspect. The company may still be waiting for inquiries. You can nevertheless establish whether the partner finished the work it reported.

For a finding, ask the partner to point to the problem. A broken form found during an audit gives the company a reason to commission a repair. At that point, the fault has been located and still needs fixing. The record should make that next assignment clear.

In a keyword plan we prepared for a recruiting-research firm in October 2026, buyers' terms had about 34,000 searches a month. That was demand in the research. The company still needed pages to address those searches, so the figure belonged with planned work. It described no traffic gained by the website.

Read a large number beside the work it describes. If it counts searches, you can discuss whether the planned pages answer them. If it measures a completed change, you can ask to see the measurement. The same question about what was counted helps you examine both kinds of claim.

Success stories hide which work was finished

A success story selects a win; it leaves you to work out what happened in the tests around it. Ask for the work behind the headline so you can see the decisions that led there.

Most growth tests do not produce a winning channel. In our October 2026 notes, we described one win in five experiments as a good outcome to aim for. That expectation reflects how much of the work involves trying ideas that may fail. A collection of wins leaves much of that work outside the picture.

The missing work matters when you are hiring someone to test several directions. You need to know how the partner handled an unfinished change, a finding that required more work, or a campaign with no response. Those situations can tell you how it decides where effort goes next.

A headline number needs its own explanation too. Search demand in a keyword plan can sound like audience growth when the surrounding story skips the distinction. Asking which kind of work produced the number makes the claim precise. You can assess the research and the proposed pages separately from any later traffic.

A promised timeframe has a similar problem. Nobody knows in advance which experiment will work, so a date for growth cannot settle whether the work deserves your trust. Ask what the partner can show at each review, including changes already made and questions still open.

Google's hiring checklist also asks whether the SEO will share the changes made to the site. That question gives you a useful way into the record. Ask to see the changes behind a success story, then follow one through to what was measured. The account becomes easier to judge when you can connect its claims to specific work.

Ask what changed after a test failed

A failed campaign becomes useful when its evidence changes the next piece of work. Ask the partner to explain the response it saw and the decision that followed.

In our own outreach, cold LinkedIn messages to 20-30 founders received no replies. We moved the effort to requests for warm introductions. The unanswered messages gave us a reason to change how we approached people. A partner describing a similar test should be able to explain that choice.

Silence alone leaves questions open. Were the people approached likely buyers? Did the message give those people a reason to respond? A partner should examine both the audience and the message before deciding what to try next. An explanation that names them gives you something more useful than a general statement that outreach failed.

Follow the answer through to the next assignment. If the audience appeared wrong, what changed about the people selected? If the message appeared wrong, what changed about the reason to reply? These questions show whether the next test addresses the problem the partner described.

Sometimes the evidence points beyond the current channel. In an October 2026 account of our work, effort moved from LinkedIn to organic search a week later. The discussion had made the search opportunity clearer, and the client wanted to pursue it. The important detail is the reason for moving the work.

Continuing a test deserves the same examination as stopping it. Ask what the partner is waiting to learn and what would make it reconsider. Its answer should connect the evidence already collected with the work it proposes to keep doing. You can then judge the reason for spending more effort there.

Judge the partner every week after you sign

After you sign, each review needs a status and decision for every test. Compare the arrangements available to you by how they handle that review, a new direction and a decision to stop.

With a first marketing hire, you train and manage the person joining your company. You help them learn about buyers and set priorities for the work. Reviewing tests belongs in that management relationship. You and the hire can look at what happened, agree the next assignment and keep a record of the decision.

An agency agreement can fix the scope at the start and quote changes separately, often with a minimum term of several months. That gives both parties a defined assignment. A good agency knows the steps needed to carry it out. Ask how its reports will show each test and how you would agree work beyond that original assignment.

With a month-to-month partner, the founder can review experiments weekly, with a status and decision for each. Research, reports and drafts sit in the company's own repository as files with history. You can return to an earlier draft or report to understand a change, and the company keeps those files when the work together ends.

Read the proposal for three things: a status and decision for every test, the terms for changing direction, and the ability to stop. Ask whether moving the work means getting another quote. Also ask what happens if the evidence says to stop while the agreement still has time to run.

Keep the research and drafts alongside the reports used in those reviews. A later meeting may send the work in another direction. With the earlier files and their history, you can see what the original decision relied on and what changed. That record lets you trace the assignments between meetings.

Ask for past results once a channel works

If a channel already brings you customers, ask candidates for results from work in that channel. You are choosing someone to develop a direction your company has already tested.

For a company still testing several channels, the immediate need is to learn which ones deserve more effort. Before buyers respond, the available evidence is the work itself and the partner's decisions. That makes completed changes, findings and the handling of failed tests useful things to inspect.

Once a channel brings customers, you can ask a narrower question. Has this specialist done work in that channel that addresses the kind of problem your company now faces? Their record becomes more relevant because you are choosing help for a direction that already has evidence behind it.

Ask the specialist to explain those results in the context of the work completed. You still need to understand what was changed and measured. The difference is that the channel's contribution to your company is already visible, so the conversation can concentrate on how to develop it. The marketing company comparison covers options for that search.

Whichever you choose, leave room for someone at your company to review and approve the work. New experiments raise questions about buyers; work in an established channel still needs your judgment about the proposed changes. Discuss that review time before signing, so the person responsible can pay attention to the work and decide what follows.

Example: sorting one client's work by kind

For a recruiting-research client, we recorded a homepage change, an indexing problem and research for future pages.

What was reportedIts kindWhat the founder could check that dayWhat still waits on buyers
Homepage load fell from 2.5 to 0.7 seconds, 3 August 2026Completed change, measuredThe before-and-after load readingsInquiries from people visiting the page
No page was in Google because the sitemap and canonical tags pointed to an old domain, 3 August 2026FindingThe domain in the sitemap and canonical tags, and the absence from GooglePeople finding the pages and responding after the search fault is addressed
The category term had 210 monthly searches; buyers' words had about 34,000, 14 October 2026Planned work: keyword planThe keyword list and its monthly search figuresBuyer visits to pages that would target those terms

A founder could inspect the homepage change and the old-domain problem when we reported them. The keyword figures described demand for future page work; putting all three in a growth headline would lose that distinction.

Five questions sort partners before you sign

  • Ask to separate past work

    Can the partner show you finished changes, separate the problems it found, and identify the work that stayed in a plan?

  • Follow a failed test

    Which test received no useful response, and what did the partner change about the recipients, wording or following assignment?

  • Check how direction changes

    How would the partner handle a decision to try another channel, including any new quote needed to move the work?

  • Establish what you keep

    What happens to the accounts, data and files when the engagement ends, and which of them will your company retain?

  • Agree your review time

    What time should you allow each week for reviewing the evidence and approving work, given the decisions the partner expects?

Sources

Put these questions to us on a call

Ask us about completed work, failed tests and how Tin decides each week what to continue, change or stop. We can talk through the questions above and the growth help your company is considering.

Book a call