Qualified demand, not raw volume

Get the Right Buyers, Not Just More Visits

Traffic only helps when the visitors fit the offer and can take the next useful step. Compare SEO, paid search, referrals, social, and email by intent, speed, control, and risk.

Start with the channel that can produce the clearest signal, then build a mix that can withstand change.
An Art Deco acquisition filter separating raw traffic from qualified buyers and measurable next steps

Define the customer before choosing the channel

Choose the next result your business needs: a booking, a suitable enquiry or a completed purchase. Then name the people who can produce it and the conditions that make them a fit.

For an upholstery workshop, that might mean people within its collection area who want a sofa recovered. A visitor looking for a free sewing pattern has a different task. Both may arrive through search; counting them together tells you little about demand for the workshop.

Write a short qualification note before starting a campaign: the work you accept, the area you cover, any minimum job requirements and the action you want the visitor to take. Use it when reviewing enquiries, not just when choosing keywords.

Match each channel to a real customer situation

Organic search

Useful question: can people find a page that answers a relevant search, such as sofa reupholstery in the workshop’s area?

Build a useful service page with examples, scope and a way to request an assessment. Follow relevant searches and suitable enquiries over time. Google’s SEO guidance says changes can take hours to months to appear, and some produce no noticeable result. Visibility is not guaranteed or permanent.

Paid search

Useful question: will a defined group of searchers request the service when the offer and destination are explicit?

Search campaigns can reach people searching for products or services. For the workshop example, keep the intended location, service and landing page aligned. Set a spending limit and inspect actual search terms and enquiry quality. A quick click report is not the same as a quick answer about profitable jobs.

Referrals

Useful question: can a relevant business or publication introduce customers whose needs fit?

A furniture repairer might refer work it does not handle to the upholsterer. Give the referring party a clear description of accepted work and a useful destination page. Record which referrals become suitable requests. One mention does not establish a dependable source of customers.

Social

Useful question: does seeing the work help someone recognise a need or evaluate the result?

Show an actual chair before and after recovering, explain the work and link to the relevant service. Distinguish people saving ideas from people asking for an assessment. Reach and reactions can describe attention; check enquiries and jobs before treating them as sales evidence.

Email and returning customers

Useful question: is there a useful reason for someone who already knows the workshop to return?

For subscribers who asked for updates, a new fabric collection could link to real examples and availability. Measure replies or requests that follow. An existing audience provides another route to customers, but a mailing list alone does not guarantee delivery, attention or repeat work.

Give the first test a question you can answer

Do not divide a small effort across five channels merely to look balanced. Choose one test based on the customer situation, the material you already have and the result you can observe.

An example test brief

  • Customer: someone in the collection area considering sofa recovering.
  • Destination: the sofa service page, with photographs and assessment requirements.
  • Action: request an assessment with the information the workshop needs.
  • Measure: suitable requests, quotes and accepted jobs, alongside visits and test cost.
  • Review: choose a date, a spending limit where applicable and a reason to pause, such as a broken form or persistent requests for work you do not offer.

Use this as a planning example. Set the actual scope and limits from your capacity and past results.

Read the result beyond the click

Bring the traffic report and the enquiry records together. A source can send many visitors while producing requests for the wrong service or location.

  1. Check arrival. Did the intended audience reach the intended page?
  2. Check action. Did the form, call or purchase route work, and was the action recorded?
  3. Check fit. Which requests met your qualification note?
  4. Check outcome. Which suitable requests became quotes, bookings or orders?
  5. Check effort. Include advertising cost and the work needed to create content or handle enquiries.

Do not add each platform’s reported conversions together and call that a count of customers. A person can encounter several channels before acting. Attribution models assign credit to interactions; they do not turn each credited interaction into a separate order. Compare reports with your own enquiry or order records.

Choose what the next channel should add

If relevant searches already produce suitable requests, a portfolio of completed work could help referred visitors assess quality. If social posts generate questions but few assessments, improve the service explanation before assuming you need another audience.

Keep the purpose of each channel explicit. The second channel might introduce new people, answer doubts or bring a previous customer back. It does not have to repeat the first channel’s job.

Review what would happen if a major source stopped sending visitors. Maintain current service pages, customer records and the relationships that support referrals. Spread effort only as far as you can keep each route useful and measure its contribution.

Bring the channel and the requests it produced

If a campaign looks busy but produces little suitable work, compare its destination page with a few recent enquiries. Note the mismatch: wrong location, wrong service, missing information or an action that fails.

Ask Devenia to review the traffic source and landing page, with the result you wanted and the result you observed.

Which source brings your most suitable requests, and what evidence connects those requests to completed work?