Waftr.com
Three colleagues reviewing a Looker Studio chart where conversions recover after a tracking fix
Case studies

Analytics case studies: what broke, what we changed, what it measured

Six analytics case studies across GA4, Tag Manager, consent and attribution, for SaaS, retail, real estate and a translation startup. Each result is published with its metric, scope, time window and the offer behind it, and clients stay anonymous unless they approved their name.

FeaturedEnterprise SaaS, server-side tagging

53% → 38%

Metric
Share of leads counted as direct
Scope
All CRM leads
Window
2 months before vs 2 months after the setup
Offer
Audit, Build, then Care
Read the featured case
Case finder

Find the analytics case study closest to your tracking problem

Filter by the symptom you see, the platform involved or the offer you are weighing. Cards that do not match drop out, and the number on each chip shows how many cases that choice would leave.

Symptom you see

Platform involved

Offer behind it

Showing 6 of 6 case studies

How we report a result

Four parts every published result carries

A percentage on its own says little. Each result here names the metric that moved, the scope it covers, the window it was measured over and the offer that produced it.

  1. 1
    The metric

    One number, stated as measured and never stretched into an "up to" range.

  2. 2
    The scope

    What the number covers: one optimized page, the rebuilt audiences, or all CRM leads.

  3. 3
    The window

    The comparison period behind it, such as 50 days before against 50 days after the form change.

  4. 4
    The offer and the client

    The Waftr offer that produced it. Clients are named only with their approval, which NexTranslate gave.

Reading a result

How to read an analytics case study: what each number covers

Scope changes what a percentage means. Three results from the finder, each read with its edges, then the three questions people ask before they trust a case study from a consultancy they have not worked with.

  • +20%
    Covers
    Remarketing conversions on the audiences rebuilt in the Build
    Outside it
    Account-wide Google Ads conversions
  • +23%
    Covers
    Conversions on one optimized page, 50 days before vs 50 days after
    Outside it
    Every other page on the site
  • 53% → 38%
    Covers
    The direct share of all CRM leads, 2 months before vs 2 months after
    Outside it
    How the 15 points split between paid search and organic

Why are most clients anonymous?

Most clients asked to stay unnamed, so their case studies name the client type instead, such as enterprise SaaS or US retail. NexTranslate is named with its approval. Named or not, each result is published the same way, with its metric, scope, time window and the offer behind it.

Are these results verified?

Each number comes from the client's analytics, ad platform or CRM data, read over the window printed beside it. NexTranslate's result is a test instead: with no choice made in the cookie banner, GA4, Google Ads, the Meta Pixel and other marketing tags were checked and found waiting for consent.

Will the same fixes work on my site?

Not by default. Each result depends on that client's traffic, stack and starting point, and none of these numbers is a forecast. What transfers is the method: find where data drops between the browser and the report, fix it at that layer, then measure the change over a stated window.

Your next step

Tell us which case looks like your setup

Pick the case closest to what you see and send your site. We compare it with that case's starting point before suggesting anything.

Your name is missing.
This email address needs a second look.
Not sure yet? Run the free Tracking Scan

What happens next: a reply within one business day comparing your setup with the case you picked.