Launch · Consulting
Landing Pages & Conversion Rate Optimisation: Message Match Before Button Colours
Most conversion problems are message problems, and most sites do not have the traffic to test their way out of them.
The problem
What this usually looks like
The traffic is arriving and the page converts at a rate nobody is happy with, so somebody suggests testing the button colour. The page opens with the company's own history, the proof sits three scrolls below the fold, the form asks for nine fields including annual revenue, and on a phone the headline is cut in half. The ad promised one specific thing and the page talks about something adjacent, so a large share of visitors leave before reading a full sentence. Then a test gets run on a few hundred visitors, produces a number, and that number gets treated as fact.
How I approach it
I start with message-to-market match, because a page that answers a different question than the ad asked cannot be fixed with layout. Then structure: the promise in the first screen, proof placed next to the claim it supports, objections handled in order, and one action instead of five. Form friction, mobile behaviour and page speed are treated as conversion factors rather than technical housekeeping, since a page that takes six seconds on a phone has already lost the visitor. Testing comes last and gets sized honestly. Most sites do not have the traffic for a statistically valid A/B test, and I will say so rather than run one anyway and report the noise as a win.
Scope
What I actually do
Check message-to-market match: what the ad, email or search result promised, and whether the first screen delivers it.
Restructure the page around promise, proof, objection handling and a single action, rather than decorating the existing order.
Place proof where the doubt happens, beside the claim, beside the price and beside the form.
Cut form friction to what the sales process genuinely needs, and move the rest into qualification after the booking.
Fix the mobile experience specifically, because that is where most of the traffic and most of the abandonment lives.
Treat page speed as a conversion factor: image weight, script bloat, Core Web Vitals and time to first meaningful content.
Instrument the page, conversion tracking, scroll depth, form field drop-off and session recordings, before changing anything.
Run a testing method sized to your actual traffic: sequential changes with qualitative evidence at low volume, controlled A/B tests only where the numbers can support them.
Deliverables
What you get
- Conversion audit of existing pages with a prioritised fix list
- Rebuilt landing pages or a page structure specification for your team to build
- Message match map pairing each traffic source with the page that answers it
- Form and qualification redesign with the fields you can safely drop
- Page speed and mobile fix list with measured before and after
- Conversion tracking and behavioural instrumentation
- Testing plan sized to your real traffic volume, including what not to test
Outcomes
What changes
- Pages that answer the question the visitor arrived with
- Fewer abandoned forms because you stopped asking for what you do not need
- Mobile stops being the version nobody checked
- Changes made on evidence instead of on whoever spoke loudest in the meeting
- An honest read on what your traffic volume can and cannot prove
Engagement
How we work together
Single page or campaign conversion audit
Landing page design and build project
Ongoing CRO retainer where traffic volume justifies continuous testing
Questions
Straight answers
Why do you push back on A/B testing?
Because the maths is unforgiving. To detect a realistic improvement you generally need hundreds of conversions per variant, and a B2B site producing thirty or forty conversions a month would need many months per test. By which time your traffic mix, your ads and your season have all changed. Most sites that run A/B tests are reading noise and acting on it confidently.
So what do we do instead at low traffic?
Fix the large, obvious defects that do not need a test to detect: message match, page structure, proof placement, form length, mobile layout and load speed. Then make one substantial change at a time, define the observation window before you start, and support it with qualitative evidence, session recordings, form field drop-off, scroll data, five to eight user tests and, best of all, listening to sales calls. Big changes read through the noise; button colours do not.
Is this design work or copywriting?
Mostly order and words. What the page says, in what sequence, with what proof beside it, and what you are asking the visitor to do. Design matters and I will build the page, but a beautifully designed page that answers the wrong question converts worse than a plain one that answers the right one.
Can you work on our existing site or does it need rebuilding?
Usually the existing site is fine to work in. I would rather restructure pages in what you already have than start a rebuild you did not ask for. If the platform itself is the constraint, unfixable load times, no ability to publish a page without a developer, no tracking. I will say so and we can treat the rebuild as a separate decision.
Related
Services that usually go with this
Corporate Website Design & 3D Web Experiences
Most corporate sites look like a template with the logo swapped. The fix is craft and a performance budget, not a bigger hero image.
Sales Funnel Design & Build
A funnel is an operations problem, not a graphics problem.
Paid Media & Performance Marketing
Paid media is an amplifier, and amplifying a funnel that leaks is an expensive way to prove it leaks.
Hiring instead of contracting? See the full-time roles, or the ATS-formatted resume.
Contact
Tell me what you are hiring for
- Call: +1 519 278 5085
- Email: tim@timarmstrong.ca
- Based: London & Southwestern Ontario, Ontario, America/Toronto (Eastern Time)
- Availability: Available nationally across Canada, on-site, hybrid or remote
- Serving: London · Stratford · Woodstock · Kitchener: Waterloo · Cambridge · Guelph · Brantford · Sarnia · Chatham-Kent · Windsor · Toronto · Mississauga and Canada-wide