Measure properly first, then add tools and AI
Most companies are not short of tools. They are short of numbers they can trust enough to judge which tool is earning its cost.
Short answer
This is a topic page, not a separate package. Recommendations on measurement, martech and AI are already part of the monthly retainer. Only a full dashboard build, a detailed tracking plan, or training your in-house team is charged as a module. MWY writes the specification and signs off the result; your agency or IT team does the implementation.
When this needs a closer look
Tracking is missing or wrong
Events dropped, pixels firing twice, Conversions API not connected, offline conversions never pushed back. Every decision downstream rests on a wrong number.
Many tools bought, few actually used
CRM, CDP, marketing automation, AI tools — paid for monthly, with nobody able to say which one changed an outcome.
You want AI, but under control
The team wants to try AI in content, advertising and customer care. You need to know where it belongs, where it does not, and how to measure either way.
Scope
MWY writes the specification and signs off the result. Implementation stays with your agency or IT team — we do not take technical build work.
- Tracking plan
- GA4, Google Tag Manager, Conversions API, offline conversions, UTM conventions
- Dashboard
- Metric framework, data sources, how to read it. Building the dashboard itself is a module
- Martech review
- What you own, what you actually use, where tools overlap, what to cut
- Tool selection
- Criteria and a shortlist. MWY sells no software and takes no vendor commission
- AI roadmap
- Where it belongs, where it does not, and how to measure the difference rather than the feeling
- Training
- A half-day workshop for your in-house team, charged as a module
Four steps
Step 1
Check the data you have
Measure the gap between platforms, GA4 and the CRM. List dropped events and double-counted ones.
Step 2
Write the tracking plan
Define each metric, its source, and who owns it. This document is the contract between marketing and IT.
Step 3
Sign off
Your agency or IT implements. MWY re-checks every event before it counts as done.
Step 4
Review tools and AI
Only once the numbers can be trusted is there any basis for judging whether a tool earns its cost.
Where we usually advise against AI
A shorter list than where it helps, and a more important one.
Regulated product claims
Cosmetics, supplements, healthcare, dentistry. One wrong efficacy claim is a compliance risk, and a model does not know where that line sits in Vietnam.
Replying to a complaint
An angry customer does not want a fast answer, they want someone accountable. Automating this breaks the relationship faster than it repairs it.
Budget allocation decisions
Automated optimisation sees only the metric it was given. It does not know your margin by product.
Bulk SEO article generation
Cheap and fast, and precisely what search engines and AI answer engines now filter out. This damages a domain rather than building one.
Brand naming and messaging
Not because a model does it badly, but because it is a decision for whoever is accountable for the brand.
Frequently asked
Is this a separate package?
No. Recommendations are already part of the monthly retainer. Only a full dashboard build, a detailed tracking plan, or training is charged as a module.
Will MWY implement GA4 and Conversions API for us?
No. We write the technical specification and sign off the result. Implementation stays with your agency or IT team. The party writing the spec and signing it off should not be the party building it.
Does MWY resell software or take vendor commission?
No, and it is written into the contract. When we recommend a tool we hand over the criteria so you can verify the choice yourself.
How long before the numbers can be trusted?
Fixing tracking usually takes a month. It then takes another quarter of clean data before comparisons hold up well enough to decide on.
Our team wants to use more AI. Will MWY block that?
We do not block it, we ask how it will be measured. An AI initiative that cannot describe its own measurement is a reason to postpone, not a reason to forbid.
Start by finding out which numbers are wrong
Month one of the retainer checks tracking and data quality across every channel.