An ad is not a recommendation.
Check the developer.
Before you commission a website, look past the promise. Open the developer’s own site, test the mobile experience and ask for evidence of the work behind the sales pitch.
Start with the checks ↓Found a web studio in an ad?
That is where the checking starts.
An advertising slot tells you that someone paid to reach an audience. It does not establish their SEO expertise, reputation or ability to deliver your project. A polished landing page is an invitation to investigate, not a reason to skip due diligence.
It also does not prove the opposite. Experienced teams advertise too. Google states that advertising does not positively or negatively affect organic search inclusion or ranking. Judge the evidence, not the acquisition channel.
Google: advertising and organic rankings ↗If a developer sells speed, their own website is a reasonable place to ask for proof.
Six checks.
Fewer expensive surprises.
You do not need to read the source code to ask good questions. Start with what a customer can see and verify.
Their own website
Try the menu, forms and service pages on your phone. Broken links, layout jumps and slow interactions deserve an explanation.
Measured performance
Test several comparable pages. Repeat lab measurements and review real-user data where available. A single score is not a verdict.
Verifiable case studies
Ask what the developer actually built, when the project was delivered and which functions you can try. A screenshot alone is not a case study.
SEO evidence
Ask for the target queries, period and starting conditions behind a result. A branded query or an undated chart does not establish broad SEO expertise.
Architecture and costs
Clarify the platform, paid components, renewals, integrations and maintenance responsibilities. Know what happens when the requirements change.
Acceptance and ownership
Agree on functionality, access to the project, backups and acceptance checks. Turn verbal promises into clear deliverables.
A weak implementation
is the warning sign.
An overloaded WordPress theme, unnecessary plugins and poorly optimized images can make a website slow. If a studio shows you that result while promising exceptional performance, ask what it will do differently for your project.
WordPress itself is not proof of low skill. Its own optimization guidance addresses themes, plugins, caching and images. The same standard applies to Django: a framework name cannot compensate for poor engineering.
My choice for custom business logic is Django and Wulver. What matters to the client is whether the agreed functions work, the site performs well and the architecture remains maintainable.
WordPress / Optimization ↗Open PageSpeed.
Then look beyond the score.
Compare the same device mode and similar page types. Keep lab results separate from field data, and do not treat missing field data as a failing result.
Main content becomes visible
The page responds to interaction
The layout remains stable
Recommended good thresholds at the 75th percentile of real visits, not test results for this website or another studio. web.dev ↗
A low result is a reason to investigate the bottleneck and ask for a plan. Repeatedly poor measurements, obvious usability failures and an unwillingness to explain them are more meaningful than the CMS logo.
Before signing,
ask these questions.
- Which live projects can I inspect, and what exactly did you implement?
- How will we measure performance, and which pages are included?
- Which features require paid licenses or recurring services?
- How will custom functions and AI integrations be maintained?
- Who owns the access, and what is handed over at completion?
- What are the acceptance criteria and the process for fixing defects?
Apply the same standard
to me.
Explore this website, check its speed and ask for relevant case studies in a private conversation. Tell me your requirements and I will explain the proposed architecture, scope and checks before we start.
Discuss the project and see cases ↗