The engagement
From first call to traffic you can grade.
Wardly is a deployed engagement rather than a signup. Here is the whole sequence, what each stage involves, and what we need from you at each point.
Scoping call
30 minutesWe map your site, your traffic and the one conversion that matters commercially. No deck and no discovery framework — the call is with the person who would do the work, and most of it is spent turning a rough answer about what your conversion is into a definition precise enough to count.
- What the site exists to produce, in commercial terms
- Whether that action is countable without a human adjudicating each case
- Roughly what volume it happens at, which determines whether ranking will be statistically meaningful
- Where it completes — on your site, or inside someone else's checkout or booking flow
Written scope
After the callYou get a document covering what we would deploy, what we need from your team, a timeline as a range rather than a promise, and the cost. Nothing is committed before you have it. If the call surfaced that this is not a fit, you get that answer instead, with the reasoning.
- The conversion event, written down as a definition
- What gets installed and how it fits your hosting
- The access and technical contact required
- Timeline range, cost, and the terms around data
Deployment and verification
Usually daysCollection is installed to run under your own setup. Then it is verified against live traffic — every path to the conversion is walked, terminal events are checked for survival across the navigation that follows them, and the numbers are reconciled against a source you already trust.
- First-party collection installed on your setup
- Every conversion path walked and confirmed to fire
- Session and visitor grouping checked across real entry points
- Reconciliation against an existing trusted source
Configuration and modelling
Days to weeksThe conversion event is wired in, behavioural scoring runs against it, and every source, channel and entry point gets ranked on that axis. Once enough conversions have accumulated, a model is trained on your own data and validated before it is allowed to score anyone. Then the reporting is built around how your team works.
- Conversion event defined and wired
- Behavioural scoring live from the first sessions
- Source ranking configured against the event
- Conversion model trained, validated, and held back if it does not clear the accuracy bar
- Reporting configured — what is surfaced, how it groups, what comes first
Ongoing review
ContinuousReporting keeps running and gets refined as the business changes. Conversion definitions move, new channels need grouping, and the ordering you thought you wanted turns out not to be the one you read. Those are configuration changes handled inside the engagement, not new projects.
- Configuration adjusted against what actually gets read
- Conversion event re-pointed when commercial priorities move
- Models retrained on a schedule as data accumulates
- New campaigns and entry points wired in
What we need from you
Less than most implementation projects, and the constraint is rarely the one people expect.
- One technical contact, for roughly an hour during deployment and reachable during verification.
- Enough access to install collection on your setup and verify it against live traffic — specified in the scope before anything starts.
- Someone who can settle what the conversion event is. This is the one that actually blocks projects.
- Where the conversion completes inside a third-party flow, somebody who knows how that integration was built.
We do not need access to your CRM, your customer database, or your payment provider.
Where projects slow down
Almost never in the technical work. Deployment and verification are the predictable part.
The delay, when there is one, is the conversion definition. Asking a business to name the single action that matters most commercially frequently surfaces that marketing and finance have been using the same word for different things — that “leads” means form submissions to one team and qualified opportunities to another, and that every report either has ever produced has been quietly answering a different question.
That conversation is worth having regardless of whether you deploy anything. It is also why we do the scoping call before quoting.
When we say no
- You want a free tool you install yourself. We do the deployment, and it is a paid engagement.
- Your site has no clear commercial conversion yet. There is nothing to rank against.
- You need procurement, SSO and a security review cycle before work can start.
- Your conversion happens entirely off-site with no digital signal at all.
Questions
Common questions.
What actually happens on the scoping call?
We look at your site and ask what it exists to produce commercially. Most of the thirty minutes goes on pinning the conversion event down to something countable — which is usually harder than people expect and is the most useful part of the call whether or not you go ahead. You leave knowing whether this is a fit.
Is there a contract before the scoping call?
No. The call is a call. The written scope comes after it, and nothing is committed until you have that in front of you with a number on it.
Who does the work?
The person on the scoping call. There is no handoff to an implementation team, because there is no implementation team — that is the whole model, and it is why we cap how many deployments run at once.
Do we have to stop using our current analytics?
No, and we would usually advise against it initially. Running both gives you a reference point while you build confidence in a new measurement layer, and it preserves any ad platform integrations that depend on your existing setup.
What if the scoping call says we're not a fit?
We tell you on the call and explain why, and where we can, we point you at what would fit better. That is a faster outcome than a proposal nobody was ever going to accept.
The first step is a thirty-minute call.
We work out what your conversion event actually is, and tell you straight whether a deployment would be worth it.