AI, Automation & Data
Data and analytics
What this category covers
Everything between raw activity and a decision somebody can defend. At one end, getting measurement onto a website correctly so the numbers mean something. At the other, warehouses, pipelines and unified customer profiles that let questions be asked across systems that were never designed to talk to each other. In between sit the services that usually come first — dashboards, data cleaning, system migrations and working out what advertising is really earning.
The work is bought for one of three reasons. Someone has lost confidence in their numbers, usually because two systems disagree and nobody can explain why. Someone is spending several days a month assembling a report by hand and cannot keep doing it. Or a decision is waiting on a question that no single system can answer, most often about which marketing genuinely produces revenue.
What separates good work here from expensive work is not the tooling. It is the discipline of defining things before building them — what a customer is, what revenue includes, what one row in a table represents — and being honest about what a dataset cannot tell you. A confident answer from unreliable inputs is the most costly output in this field, because it looks exactly like a reliable one and gets acted on for months.
- Services
- 10
- Detailed pages
- 10
10
Data & Analytics
10 services
-
01
Analytics implementation (GA4 / GTM)
Events, goals and e-commerce tracking — data worth trusting.
-
02
BI dashboards
Looker, Power BI or Metabase — views you can decide from.
-
03
Data warehouse setup
A single source of truth in BigQuery or Snowflake.
-
04
ETL / data pipelines
Data pulled from systems, cleaned, and scheduled.
-
05
Data cleaning & migration
Duplicates, formats and legacy migrations put right.
-
06
Customer data platform setup
Unified customer profiles across tools and channels.
-
07
Attribution modelling
Which channel is actually bringing the revenue.
-
08
Automated KPI reporting
Weekly or monthly packs that build and deliver themselves.
-
09
Data strategy & governance
Ownership, quality rules, access and retention policy.
-
10
Survey & market research analysis
Collection, statistical analysis and an insight report.
How the work runs.
-
01
Scope
What you actually need from Data & Analytics, written as deliverables rather than adjectives.
A day or two
-
02
Price and date
A fixed scope, a number and a delivery date, agreed before anyone starts.
Same week
-
03
Build
The work itself, with something you can open and look at rather than a status report.
The bulk of it
-
04
Handover
Files, access and a walkthrough. Nothing is held hostage.
Final day
See where the site stands before you brief anyone.
Before scoping Data & Analytics work, it helps to know what the site is doing today. The audit reports what it checked, what it found, and — the part most tools leave out — what it could not check, and why.
Questions people ask before they commit.
- We know our reporting is a mess but not where to start. What comes first?
- An audit, and it is deliberately a small paid piece of work rather than a free sales exercise. Two or three pages stating what is currently installed, what is broken, and which of your existing numbers can and cannot be relied on. It costs little, it takes days rather than weeks, and it means the larger decision is made with evidence. It also protects you from paying for a dashboard or an attribution model built on inputs that were never going to support it.
- Why do our systems all give different numbers for the same thing?
- Usually for one of three reasons, and they need separating before anything else. The systems are counting genuinely different things, which is normal and can be documented. The definitions differ — one includes tax or shipping, another treats a refund as belonging to the month it was issued. Or something is actually broken, such as a tag installed twice or a filter that was created and never activated. The first two are explained and written down. Only the third is repaired.
- Can historical data be corrected?
- In analytics platforms, generally no — data is written once at collection and stays as it was recorded. In your own databases and CRM, often yes, which is what cleaning and migration work exists to do. We will always tell you which category your situation falls into at the start rather than letting you assume the past can be fixed, because plans built on that assumption tend to fail late.
- Do we need a warehouse, or is that overkill for us?
- You may well not need one, and saying so costs us a larger project. A warehouse earns its cost when the questions genuinely sit between systems, or when reporting is consuming days of skilled time each month. If your questions can be answered inside one platform, or by joining two sources in a reporting tool, that is the better answer and it is what we will propose.
- How is our data handled while you are working on it?
- On accounts created for your project or inside your own environment, never in anyone's personal storage. Working copies are deleted at the end and you are told in writing that they have been. Personal data is never sent into analytics or advertising tools, because doing so breaches their policies and can result in a property being terminated along with its history.
- What will you not promise?
- That the numbers will be complete. Web measurement loses a share of activity to blockers, declined cookies, device privacy settings and people using more than one device, and any supplier claiming otherwise is either uninformed or selling. Analytics figures are dependable for trends and directional decisions. For the final financial account, your payment and accounting systems are authoritative, and we will document the expected gap between the two rather than pretending it does not exist.