The agency owner’s field guide
Find companies worth talking to. Give yourself a reason to talk to them.
Download the editable worksheet (Markdown) →
Use the directory to find companies that appear to buy work related to yours, then investigate whether there is a sensible reason to speak with them. Finish with a small, explained list and a useful question for each company.
A spreadsheet full of logos is an excellent way to feel busy. The valuable part is understanding why a company might need your particular help and what you still need to find out.
Use this when: you have a reasonably specific offer and want to research a manageable batch of potential clients. If your offer is “marketing for anyone with a budget,” narrow it before starting.
Your output: five companies to investigate further, each with a source, a fit hypothesis, an unknown and a next action. Five is a useful first-session limit, not a promise that every search will produce five qualified prospects.
1. Define a buyer you can actually help
Write one sentence naming the business, the problem and the conditions that make your work useful. For example: “We help established skincare brands improve repeat purchasing through email and lifecycle programs, when they have enough customer history and someone who can implement changes.”
Then name three reasons you would decline the work. Perhaps the business is pre-launch, needs a service you do not deliver, or cannot supply the internal resources your process requires. These are research instructions. They keep a familiar brand name from overriding your judgment.
Separate what public research can establish from what a conversation must confirm. A product range and distribution model may be visible. Available budget, margin constraints, readiness and decision authority usually need direct discussion. Record unknowns as unknowns.
2. Find agencies that serve a similar buyer
Open Browse Agencies. Search the service and industry you care about, then use the available filters. Try a specific combination such as beauty email marketing. If results are thin, remove one constraint and inspect the broader group.
Read a handful of promising profiles. Look at the clients they name, case studies and links to their own websites. Choose agencies because their work resembles what you sell, not because they rank first. Directory order is not an endorsement or a complete map of your market.
Collect possible client companies from the relevant work. An agency may offer ten services but have sold only one to a particular client. A client logo beside a service menu does not establish which service was bought.
3. Follow the company back to the work
Look up a candidate in Brands. Select the matching company, open agency profiles and follow “Named on…” links to the original sources. Read what each source actually describes. Check trading names and sub-brands if needed, keeping different businesses and locations separate.
For each useful connection, record the client, agency, source URL, date you checked it, project dates if stated, work described and agency role. A client roster may establish only that the agency names the company. A detailed case study can support a more specific buying hypothesis.
Do not treat several agency mentions as evidence of churn. The agencies may handle different markets or disciplines, or their work may be years apart. Missing coverage also does not mean the company has never bought outside help.
4. Add a current reason to investigate
Visit the company's own website and inspect relevant public announcements or current activity. Look for a development that could make your expertise useful: a new product category, expansion into another market, a changed customer journey or a publicly described operational priority. Save the source and its date.
A development creates a question. It does not establish pain. A new product line might increase the complexity of lifecycle communications, but the company may already have that covered. Your job is to formulate a reasonable hypothesis and test it.
- Investigate now: the business fits, the evidence relates to your work and there is a specific question worth pursuing.
- Research further: the business looks relevant, but scope, timing or identity remains unclear.
- Set aside: a known mismatch outweighs the appeal of the logo, or you cannot explain why a conversation would help.
Avoid a numerical “buying intent” score based on client mentions. It turns uncertainty into a very convincing-looking number.
Worked example: from a case study to a useful question
Illustrative example: the companies and findings below are fictional. They show the reasoning process, not actual directory records.
You run a lifecycle agency for skincare businesses. In an agency profile, you find a case study about a fictional company called Alder Skin. The original source describes a storefront redesign completed two years ago. Alder's own recent announcement describes a new refill product range.
Supported: the agency reports a storefront project, and the company announced a refill range. Your hypothesis: different replenishment cycles may require different customer messaging. Unknown: how Alder handles that today, whether it wants help and which team owns the work.
You do not label Alder “unhappy with its web agency” or “needs email.” You move it into the investigate group and prepare this question: “With the refill range, how are you handling reminders for customers whose products run out at different times?”
That question has a connection to the business. It also allows the answer to be “we already do that well,” which is useful qualification.
5. Make the first conversation earn its place
Turn your strongest finding into a brief opening. Mention the current development, ask about the relevant decision, and offer a small piece of useful thinking you can actually supply. Keep the agency relationship research as context unless discussing it helps the recipient.
“I saw the refill launch. How are you handling replenishment timing across the different products? We work on that problem for skincare teams. If useful, I can send three questions we use to check whether the reminders match how customers actually use the product.”
Adapt the wording to the evidence you found. Do not claim to know their internal numbers or manufacture a relationship with an agency you discovered. If they respond, qualify the problem and current ownership before discussing a solution.
The goal of the research is a better conversation. A company appearing in a directory is not an invitation to add it to an automated sequence.
Finish the session with decisions
Use the worksheet in your own document editor. Give each candidate a next action, owner and date. Stop when you have a small batch you can responsibly investigate; collecting another hundred names will not answer the questions about the first five.
After conversations, record which assumptions were right and which were wrong. If the same mismatch repeats, change your selection criteria. Evaluate the usefulness of the conversations and resulting opportunities, not just the number of companies added.
Research your competitive set if you need better agency starting points. Use the brand-page guide for name variants and evidence interpretation.