Website study / An example for Independent furniture
Alder & Form
How do you help somebody choose a chair they cannot yet sit in?
A beautiful chair earns attention. A useful buying experience also answers the awkward questions: will it fit, how does the fabric feel, and what happens if the finish is wrong for the room? This concept turns those uncertainties into a practical research and software brief.
- The buyer
- Picture the room
- The website
- Compare the detail
- The working side
- Prepare the order
Connection diagram · proposed system, not measured results
01 / Behind the website
The chair is visible. The uncertainty is less obvious.
A household buyer, an interior designer and a purchasing colleague can admire the same photograph while needing different facts. The research would separate those decisions before writing the product page.
- 01
The room planner
Wants reliable dimensions, room context and material choices without guessing from a hero photograph.
- 02
The trade buyer
May need quantities, specification sheets and a shareable finish reference before seeking approval.
- 03
The fulfilment colleague
Needs one agreed configuration, delivery requirements and a clear record of what was promised.
The uncertainty to remove
Showing a material beside its name can help people recognise the option instead of remembering it from another screen. That is a usability principle to test, not a claim that sage upholstery makes everybody buy.
A restrained opening leaves room for the product, but restraint must not hide practical answers. We would investigate which concerns actually delay a decision, including maintenance, access, returns and the difference between a sample and a final finish.
02 / Behind the website
Every visible choice should earn its place.
The supplied composition already contains a product, material swatches, dimensions and room context. Each element becomes a specific proposition rather than decoration added after the copy.
- 01
One chair, enough context
The large chair picture establishes shape and character; the adjacent sentence states the category and use. A real project would use accurate product photography and accessible text equivalents.
Check it: Ask a participant to describe the product and its intended use without prompting; record mistaken assumptions.
- 02
Materials beside the choice
Oak and sage swatches make two important attributes visible together. Detailed fabric composition, care and finish variation would belong to the related product information.
Check it: Give a buyer two requirements and ask them to identify the matching configuration and the remaining uncertainty.
- 03
Dimensions at the decision point
A dimension drawing and a room photograph can answer different scale questions. Neither should silently stand in for certified measurements.
Check it: Ask a participant to check a stated doorway and room size; count whether they use the correct dimensions.
Method behind these choices: How People Read Online: New and Old Findings (source, new tab).
03 / Behind the website
Observe the decision before polishing the product page.
The research would follow a real furniture selection, including the email or document that a buyer sends to somebody else for approval.
- Listen to the last purchase
Interview prospective retail and trade buyers about a recent choice. Ask what they compared, which information was missing and what caused a return or delay.
- Build the uncertainty map
Group actual questions into fit, feel, finish, service and fulfilment. Keep unsupported team guesses in a separate assumptions list.
- Try two information orders
Prototype a product-led opening and a specification-led route. Reuse the same accurate facts so the comparison tests organisation rather than unequal evidence.
- Watch a realistic decision
Use room dimensions and a short brief. Let participants work without coaching, then ask what they would need before requesting a sample or quote.
Three realistic tasks for a prototype
Choose a plausible configuration
Find a material and size that meet the supplied room brief. Explain the choice in your own words.
Record correct identification, elapsed time and unresolved questions.Prepare a shareable shortlist
Find enough information to explain this chair to an approving colleague who has not seen the page.
Check whether the shared information preserves finish and size identifiers.Ask a useful question
Request a sample or quotation without accidentally committing to an order.
Record completion and whether the buyer understands what happens next.
What would make this an honest evaluation?
Use the same room brief when comparing versions, alternate the order where possible and avoid describing the newer version as the better one. Separate first-time visitors from experienced trade purchasers.
A usability round can reveal a confusing label without estimating a population conversion rate. For a quantitative comparison, define the primary measure, practical improvement worth detecting, recruitment and sample size before collecting results. Report uncertainty and abandoned tasks.
Research practice: Using moderated usability testing (source, new tab).
04 / Behind the website
Original words. A reason for every page.
The writing would start with actual product facts and the buyer’s vocabulary. A reusable product system can still contain original descriptions for genuinely different products; custom does not mean manually duplicating a page for every finish.
| Question or search intent | Original page response | Evidence to gather first |
|---|---|---|
| “oak dining chair dimensions” | A product page with legible size drawing, units, materials and accessible specifications. | Approved measurements, product photographs and the manufacturer’s verified specification. |
| “sage upholstered chair care” | A useful care section explaining cleaning, fabric composition and limitations. | Fabric supplier instructions, relevant restrictions and ownership of updates. |
| “furniture trade quotation” | A clear trade route with quantity, configuration and delivery questions. | The actual approval process, service area and information needed to quote. |
The content plan would connect collection, product, care and trade pages with descriptive links. Search titles would explain the specific page rather than promise an unsupported “best chair” ranking.
Before publication, somebody responsible for the furniture would approve dimensions and claims. Search Console could later show query and indexing patterns; enquiries and order records would answer different questions about commercial usefulness.
Search phrases here are research seeds, not verified search volumes, rankings or traffic forecasts. Keyword research would validate demand, location, language and competition before publication.
05 / Behind the website
Make the assumptions visible.
A funnel makes the dependencies visible. This example uses invented inputs so you can see why a larger visitor count is not the same thing as a profitable order book.
Change the assumptions
From relevant visits to indicative contribution
Change one assumption at a time. Percentages are conditional on the preceding stage.
Illustrative calculation · not a forecast
- Expected requests / month
- 48
- Expected orders / month
- 9.6
- Indicative contribution / month
- £1,728
The arithmetic, openly shown
Requests = relevant visits × request rate ÷ 100Orders = requests × order rate ÷ 100Contribution = orders × contribution per order
An expected 9.6 orders is a mathematical average, not a promise or a fractional customer. This is contribution before fixed overheads, site cost, acquisition cost and any omitted returns or fulfilment costs.
From an attractive calculation to a trustworthy decision
Measure visits, valid requests and confirmed orders within a consistent cohort and attribution window. Deduplicate repeated sample requests and exclude staff traffic before calculating rates.
A sensitivity check matters more than the impressive-looking headline: halving the assumed order rate halves this model’s contribution. Compare that exposure with the cost of research and delivery; do not call the difference a measured return on investment.
06 / Behind the website
A product story in front. A consistent configuration behind.
Possible software could preserve the exact product configuration as a person moves from browsing to a sample request and then to a human-approved quote.
- 01
Configuration record
Store the chosen size, finish and upholstery against a versioned product rather than a free-text email.
Information in: Product ID, variant IDs and chosen quantity.
Useful output: A stable configuration that buyer and team can both check. - 02
Sample request desk
Track sample dispatch, follow-up preferences and unanswered questions without treating a sample as a sale.
Information in: Contact consent, selected materials and delivery details.
Useful output: A review queue and a clear status for each request. - 03
Quotation workspace
Let authorised staff add quantities, lead-time assumptions and approved commercial terms.
Information in: Versioned configuration, delivery constraints and current price rules.
Useful output: A quote with an expiry, named reviewer and revision history. - 04
Product content governance
Keep dimensions, care instructions and specification files connected to the correct model and revision.
Information in: Approved technical facts and a content owner.
Useful output: Consistent public information with a traceable update record.
Permissions, human decisions and things that can go wrong
A public visitor should never see another buyer’s quote. Trade prices, addresses and negotiation notes would need role-based permissions and retention rules. Payment, if commissioned, should use an appropriate payment provider rather than storing card details here.
If a finish is discontinued after a request, the software should flag the conflict for staff. It should not silently replace the chosen material or send an automatically invented delivery promise.
Potential software only. These modules are not running behind the pictured website. Requirements, integrations, data protection, capacity, cost and support would be agreed before a real build.
07 / Behind the website
Evidence informs the idea. Testing earns confidence.
The reading below supports a way of investigating product decisions. It does not prove that this imagined chair store has customers, stock, software or a measured result.
- Keep the specification correct
Agree the source of product truth and test the effect of an updated measurement or withdrawn material.
- Measure comprehension first
Treat correct configuration and understanding of the next step as separate measures from revenue.
- Test the less convenient conditions
Check keyboard use, zoom, small screens, missing images and a request interrupted halfway through.
Read the sources and the limits of each one
Learning about users and their needs (source, new tab)
Investigate users’ goals, current approach and difficulties. Treat untested internal opinions as assumptions.
Scope: Public-service research guidance includes people who use, provide and support a service.
Limit: Transferable practice guidance, not a B2B revenue experiment or a claim that research has already been conducted for these concepts.
How People Read Online: New and Old Findings (source, new tab)
People commonly scan for information, with reading depth varying by task, motivation, focus and person.
Scope: The publisher summarises eyetracking research spanning 13 years, more than 500 participants and over 750 hours.
Limit: Reading is not always shallow and no single scanning pattern fits every page. This is not a measured sales uplift for these concepts.
Memory Recognition and Recall in User Interfaces (source, new tab)
Visible choices and contextual cues can support recognition instead of asking people to recall information unaided.
Scope: A cognitive-psychology explanation applied to menus, search and interface content; published January 2024.
Limit: A design heuristic, not proof of universal behaviour or a specific conversion increase. Labels still need testing with the intended audience.
Using moderated usability testing (source, new tab)
Observe actual or likely users attempting believable tasks to reveal language, navigation and interaction problems.
Scope: Guidance covers recruitment, neutral task instructions, consistent discussion guides and accessible sessions.
Limit: Small qualitative tests expose contextual problems; they do not establish population conversion rates or commercial causation.
Creating helpful, reliable, people-first content (source, new tab)
Create original, useful content for an intended audience, drawing on genuine knowledge and clear sourcing; apply SEO to help that content be discovered.
Scope: Google’s published content self-assessment guidance, including purpose, originality, accuracy and trust.
Limit: Guidance is not a ranking guarantee. Length, exaggerated claims and copied summaries do not substitute for useful information.
SEO Starter Guide (source, new tab)
Write useful original content, anticipate the words different readers search for, and connect related pages with descriptive links.
Scope: Official introductory guidance on content, discoverability, images, search titles, snippets and site organisation.
Limit: Google does not guarantee indexing or first place. Search snippets may come from page text rather than the meta description; keyword repetition is not a sound strategy.
Involving Users in Evaluating Web Accessibility (source, new tab)
Working with disabled users reveals practical barriers; combine it with standards-based accessibility evaluation.
Scope: Guidance covers user evaluation, assistive technology, reporting scope and cautious interpretation of small studies.
Limit: User testing alone does not establish WCAG conformance, and a few participants cannot represent every disability or access need.
Web Vitals (source, new tab)
The published good-experience thresholds are LCP at most 2.5 seconds, INP at most 200 milliseconds and CLS at most 0.1.
Scope: Assess the 75th percentile of real page loads, separately for mobile and desktop, across all three Core Web Vitals.
Limit: These are performance targets, not measured results for the concepts. A laboratory result is diagnostic and is not equivalent to representative field data or a sales forecast.
Your business would have its own starting point
Start with the question your buyer keeps asking.
Bring Darren a product, a recurring enquiry or an awkward handover. That gives the first research conversation something concrete to investigate.