Judge demand from searches by separating research from buying intent, checking local and urgent modifiers, and comparing qualified visits with paid sales.
Short answer: Buying intent predicts demand better than raw search volume. Prioritise queries that include a service, location, price, provider, availability, deadline or comparison close to purchase, even when their reported volume is small. Use volume only after classifying intent and checking whether your contribution can support acquisition. The final evidence is enquiries and paid sales from qualified searchers, not an estimated monthly query count.
A popular search can describe curiosity, education, employment, entertainment or a do-it-yourself alternative. None requires a purchase. A low-volume search such as “same-day sash cord repair Bristol” can contain more buyers per search because the problem, location and timing are explicit.
Search data is useful, but it measures language and activity before it measures your business. You must connect the query to a decision and an economically viable offer.
Use the Query Commitment Spectrum
The Query Commitment Spectrum classifies searches by the next action the person is likely to take. Read the actual results as well as the words because the same phrase can carry different intent in different markets.
- Learn: Query pattern: “how”, “why” or a broad topic; Likely need: Information or self-help; Demand value: Useful for language, weak purchase evidence
- Explore: Query pattern: Category, examples or ideas; Likely need: Understand available approaches; Demand value: Early commercial possibility
- Compare: Query pattern: “best”, “versus”, reviews or cost; Likely need: Evaluate alternatives; Demand value: Stronger if your offer fits the comparison
- Source: Query pattern: Service plus location, provider or quote; Likely need: Find someone able to deliver; Demand value: High buying relevance
- Act: Query pattern: Urgent timing, availability or booking; Likely need: Resolve the problem now; Demand value: Strong intent, often lower volume
Do not assign intent from one modifier alone. “Best accounting jobs” and “best accountant near me” both contain “best” but support different businesses. Open the results and identify what search engines believe users want.
Start with the buying event and work backwards
Describe the moment a customer decides to act. What happened, where are they, how soon do they need the result and what would they search if they did not know your business?
Create query groups for the problem, service, location, price, urgency and alternatives. A customer may search the symptom before learning the service name. That query can still carry intent if the consequence is immediate.
My view is that a first-time founder should prefer 20 reachable, high-intent searches over 2,000 broad informational searches. The smaller number can support a focused local service. The larger number may require months of publishing, filtering and follow-up before producing the same sale.
This is not an argument against informational demand. It can support trust and future discovery. It is an argument against using attention as evidence that your initial offer will sell.
Treat volume estimates as ranges
Reported search volume is an estimate based on a data source, geography, device mix and period. Similar phrases may be grouped, and rare local queries may appear as zero despite real searches. Seasonality and events also move the number.
Record the source, country, date and whether the figure is an average. Use low and high cases rather than one exact count. Check several related queries and the visible activity of competitors.
Do not multiply volume by an assumed conversion rate taken from another industry. Your ranking, location, price, reputation and page determine what portion you can reach. A query estimate describes the whole search environment, not your traffic.
When no volume appears, look for completed customer behaviour: calls, marketplace requests, competitor reviews, local recommendations and paid tests. Search is one demand signal, not the market itself.
Examine the alternatives in the results
Search results reveal what the customer considers substitutable. You may find do-it-yourself instructions, directories, established providers, retailers or a different solution entirely.
Classify the first page by result type:
- Guides and videos: What it may mean: Searchers often want to solve it themselves; Question for your idea: Is the paid consequence high enough?
- Local providers: What it may mean: Commercial demand is recognised; Question for your idea: Can you meet location, trust and response standards?
- Retail products: What it may mean: Buyers may prefer ownership to service; Question for your idea: Does your offer solve a different constraint?
- Forums and unanswered questions: What it may mean: Language is visible but provision may be weak; Question for your idea: Is there budget or only frustration?
Competition in results is not automatically bad. Providers paying for visibility or maintaining current pages can indicate a valuable query. You still need to test whether your reachable share and contribution work.
Connect queries to contribution
Track the full path from query group to qualified enquiry, sale and contribution. Grouping matters. Do not compare all search visits with total sales if some traffic arrives for unrelated advice.
Use: Visit-to-sale rate = paid sales from a query group divided by visits from that group
Contribution from a query group = paid sales × contribution per sale minus query-specific acquisition cost
Include your time creating and maintaining the route. Organic visibility may avoid a payment per click but still consume founder hours. Paid traffic can produce faster evidence but should have a fixed cash limit and conversion test.
Worked example: Mae's sash-window repair service
Mae runs a sash-window repair service. Over a 30-day test, an informational page answering “how to stop a sash window rattling” receives 400 relevant search visits. It produces four qualified enquiries and one £280 job.
A service page focused on repair, location and booking receives only 75 relevant search visits. It produces nine qualified enquiries and four £280 jobs.
Visit-to-sale rates are:
- Informational: 1 divided by 400 = 0.25 per cent
- High intent: 4 divided by 75 = 5.33 per cent
Materials, travel and sale-specific costs are £90 per job, so contribution is: £280 minus £90 = £190.
The informational group produces 1 × £190 = £190 contribution. The high-intent group produces 4 × £190 = £760, despite having 325 fewer visits.
The example does not prove that every high-intent page will perform this way. It shows why volume alone would have led Mae to value the wrong demand signal. She should investigate which service queries produced the four jobs, whether the rate repeats and how much time or cash each page required.
Work in customers' properties can carry insurance, safety, consumer and trade requirements. These vary by location and service, so Mae should check current official guidance and qualified local advice.
Look beyond exact wording
Customers use different words for the same buying event. Group phrases by intent and problem rather than treating every variation as a separate market. Spoken searches can be longer, while returning customers may search a business name.
Also distinguish searches for employment, training, free instructions and definitions. A phrase containing your service name may still come from someone who wants to do the work themselves.
Talk to recent buyers and ask what they searched, what results they considered and what caused action. Their recall is imperfect, so compare it with your own enquiry and referral records where available.
Related guides
Test search demand over 14 days
Start with ten phrases across the Query Commitment Spectrum. Inspect results in the relevant location and classify the likely next action. Choose three high-intent groups connected to one offer.
Then act in this order:
- State the offer, price range, geography and response window clearly.
- Create one credible route for each selected intent group to enquire.
- Record qualified visits, enquiries, paid sales and contribution separately.
- Compare actual rates with your required sales, not with traffic ambitions.
- Continue only where searchers reach a buying decision at affordable effort.
If no sales appear, diagnose visibility, intent, offer and trust separately. More volume is not the first remedy.
Frequently asked questions
Does high search volume prove there is demand?
No. It proves that people search the phrase or a grouped variation at an estimated frequency. They may want free information, employment, entertainment or an alternative to buying. Classify the results and follow the query through qualified enquiries to paid sales.
High volume becomes commercially useful when the buyer intent, reachable visibility and contribution work together. The exception is an established transaction query with clear provider results, but your new offer still must earn trust and a viable share. Treat every published volume as indicative and check the relevant geography and period.
Can a business work when search volume appears to be zero?
Yes. Data sources may suppress very small local counts, group wording or miss demand created through referrals, physical locations and outbound sales. Some customers do not know the service category until a provider explains it. Look for current workarounds, competitor transactions, local requests and direct paid tests.
Zero reported volume means search may not be the primary route, not that no buyer exists. The exception is a model that depends entirely on customers independently searching that exact category. It then needs stronger evidence that enough queries occur despite the missing estimate.
Should I target informational searches at all?
Yes, when they come from your intended buyer and naturally precede a purchase you can serve. Informational material can establish competence and help customers diagnose a problem. Calculate the time to create it and track assisted enquiries rather than assuming every reader becomes a lead.
Prioritise high-intent coverage first when you need early sales. The exception is a new or complex category where buyers require substantial education before they can compare providers. Even then, include a clear commercial next step and measure whether education advances real decisions.
Are “near me” searches always high intent?
They are often commercially relevant because location matters, but not always. A person may be comparing options, checking opening times or researching without an immediate need. Inspect the results, device context and action taken. For a local business, verify whether the searcher's location falls inside a profitable service area.
Calls from outside your route add volume without usable demand. The exception is a destination business whose customers travel, where a wider catchment may be valuable. Test actual origin, purchase and contribution rather than assuming the modifier guarantees a sale.
How should I compare paid and unpaid search traffic?
Compare contribution after the complete acquisition cost and founder time. Paid search provides controlled exposure and faster data, but every click consumes cash. Unpaid visibility avoids a click charge yet requires creation, maintenance and time before results. Group both by intent and track qualified enquiries, sales and contribution.
Do not compare cost per visit alone. The exception is an early test designed only to learn which message attracts qualified buyers. Set a fixed learning budget and do not call clicks profitable demand until sales economics are known.
How often should I recheck search demand?
Recheck when season, geography, customer language, competition or your offer changes, and compare like periods. A quarterly review can be a practical working rhythm for an active small business, while strongly seasonal markets need closer tracking around key buying windows.
Do not react to one week's fluctuation. Keep the original source and date so changes are comparable. The exception is a sudden event, regulation or platform change that alters the buying mechanism. Investigate immediately, then confirm whether the shift persists before rebuilding the business around it.
Comments (0)