The claim: “We put you at number one”
The first question is number one for what. Ranking for an invented phrase, a long branded sentence or a service nobody buys can produce an attractive screenshot without establishing a meaningful acquisition channel. The ranking and its commercial interpretation are separate claims.
At the same time, dismissing every low-volume term would be poor B2B judgment. An engineering buyer searching for an unusual tolerance, material or approval requirement may be far more relevant than thousands of general readers. The useful distinction is not simply high volume versus low volume. It is evidenced buyer relevance versus a phrase selected mainly because it is easy to win. Google Ads: keyword estimates and competition definitions.
Four different numbers that should not be called demand
| Measure | What it describes | What it does not establish |
|---|---|---|
| Estimated search volume | A provider's modeled historical search activity for a defined market | Future visits to your page |
| Search Console impressions | Recorded appearances of your property under that report's definitions | The total market demand you could capture |
| Organic clicks | Clicks attributed to search results in that source | Completed inquiries or new customers |
| Accepted inquiries | Requests that meet your documented intake definition | Closed sales or profitable retention |
Keyword-research metrics need a definition, market and collection date. A missing record or null value should stay unknown; it should not be silently relabeled as proof that nobody searches.
A practical query-value review
Start by writing the buyer problem in normal language. Identify which offering solves it, who can purchase that offering and where the business can serve them. If a keyword cannot be connected to those facts, a ranking target needs a stronger rationale.
Next, inspect the live result set. Are the ranking pages relevant suppliers, technical references, unrelated meanings or mostly the provider's own assets? A keyword-difficulty estimate is an aid to investigation, not a substitute for this inspection. Paid-ad competition is also a different concept from difficulty in organic results.
Finally, look for independent evidence of the question: actual sales questions, public procurement requirements, relevant Search Console queries, useful site-search terms or supportable customer research. Keep the provenance of each observation. A phrase suggested by an AI model is a candidate to validate, not independent demand evidence. Google: Search Console performance report.
Original sensitivity example: small can matter, but the assumptions must stay visible
Suppose a hypothetical niche page receives 20 relevant visits in a month. If one in 20 requests a consultation, that produces one inquiry. If one in four suitable inquiries later becomes a customer, the scenario averages one customer per four comparable months. These are illustrative assumptions, not ZeroDark benchmarks or forecasts.
Change the assumptions and the economics change. If most visitors are students, the inquiry rate may be much lower. If the offer is unavailable in their region, the apparent interest may have no practical value. A provider claiming commercial success should explain these constraints rather than hide behind position alone.
Spot query selection that makes a guarantee easier
Look for a contract that allows the provider to choose or substitute terms without your approval. Check whether it counts your business name, spelling variants and near-identical phrases as multiple wins. Ask whether the reporting excludes terms with weak results or replaces them after each measurement.
Those practices are not automatically evidence of wrongdoing. They are reasons to inspect the definition of success. A defensible scope states the queries or selection criteria, relevance rationale, market, measurement method and rules for changing the set.
What should the report show?
Keep a query register with the proposed phrase, buyer problem, supporting demand evidence, provider estimate, observed impressions and destination page. Add a separate field for relevant inquiries. Mark unknowns explicitly. This makes it possible to distinguish a retrieval milestone from an acquisition result without pretending every valuable B2B topic has a large tool estimate.
ZeroDark's position is straightforward: a technically real ranking deserves an accurate label. Commercial SEO value requires additional evidence. The provider should welcome that distinction because it makes useful work easier to defend.
Supporting documentation
Google Ads: keyword estimates and competition definitionsGoogle: Search Console performance reportGoogle: hiring an SEO