How it works
Fettle goes on top. Nothing gets replaced.
Whatever runs your warehouse, and however it was set up at go-live, Fettle works with what's already there.
Fettle reads the WMS, the other systems your warehouse runs on and whatever records who clocked on and when, then works with the data as it stands. It doesn't depend on the WMS having been set up for the questions you're asking now, since most were set up to run the warehouse rather than to explain it.
Before
The answers are in your systems, out of reach
The data has sat in your systems for years. Every system keeps it its own way, so questions get answered by hand, late or not at all.
- GMWhere are we against plan?
- OpsWhat's happening on the floor?
- PlanningCan we meet demand?
- FinanceWhat's the real cost?
- Reports built by hand
- Figures that don't agree
- Work the WMS never sees
- Answers a month late
- Spreadsheets in between
- Questions nobody asks
The systems you run
- CRM
- OMS
- ERP
- WMS
- Returns
- Data warehouse
Six systems, each keeping the data its own way
After
Each manager opens on today's question
Fettle goes on top of the systems you run and reads them every day, so each manager sees what needs attention in the part of the warehouse they answer for.
- GMWill we deliver today's commitments?
- OpsAre we using our people effectively?
- PlanningIs the warehouse set up to work efficiently?
- FinanceAre we making money on the work we do?
Operational views for each role The command view shows what needs attention today
Your data, read by Fettle
- Nothing gets replaced
- Two or more days on site in week one
- Views configured within four weeks
The same systems
- CRM
- OMS
- ERP
- WMS
- Returns
- Data warehouse
The same six systems, read every day
Protect the core, close the gaps
The systems you've invested in carry on doing the job they're good at while Fettle closes the gaps between them. What gets built on top stays yours: change the WMS underneath, the ERP or the vendor, and not one operational view, KPI or year of history goes with it. Your own team can query your data with your own tools whenever they want.
Our founder spent twenty-eight years as an IT Director living with every system he bought, which is where this way of working comes from. A replacement often shows its gaps only after a long and costly implementation, by which point the business has paid for the risk and still doesn't have quite the fit it needed. Building across the systems you already have gets to the value sooner and uses the money already spent rather than writing it off.
Could we just build this ourselves?
You could build the data side, with your team keeping full access to the data whichever way you go. What takes longer is knowing which questions matter on a warehouse floor and how to measure people fairly, which comes from twenty-eight years on the warehouse side rather than from a data model. Fettle gets you there in weeks.
Is it safe to run AI on our operational data?
It is when three things hold: the AI reads your data only after it's been standardised and checked, the data stays yours and the company handling it is certified to ISO 27001. Fettle works that way. AI is only as good as the data going into it, so nothing reads the raw feed from each system.
A lot of software reads whatever data is easiest to reach. Fettle reads what matters on the floor, because we already know what that is: where each location sits, how fast each product moves, what a fair time for a task looks like and the service rules for each kind of order. That knowledge goes in when your data is standardised and checked, before anything reads it, including Albie, the assistant that answers questions about your operation in plain English.
Why this couldn't be done before
The data has sat in your systems for years. Reaching it was hard, because every system keeps it its own way and building the software to join it up cost more than one warehouse could justify. AI changed both halves, cutting the cost of writing the software that gets at the data and pulling the answer out once it's there.
ONE WEEK, ONE WAREHOUSE
Time between clocking on and the first task
572 hours the same as the whole paid week of more than fourteen people on 40-hour contracts
572 hours in one week between operators clocking on and starting their first task, drawn as fourteen full 40-hour weeks and part of a fifteenth.
One week in a warehouse we've worked in.
How long before we see anything?
You'll see your first numbers within two weeks, so we can validate them together, and have operational views configured to your warehouse by week four. Most early gains come from improving how your warehouse already runs, so nothing needs to be rebuilt or reorganised to reach them.
We spend two or more days on site in the first week. In that time we map the layout, walk a day in the life of the work with the people who do it, write down the service level for each kind of order and agree the starting point every change gets measured against.
Your own rules go into Fettle from that week: the service level for each kind of order, how your shifts and teams are set up and your local exceptions, so the operation doesn't have to work round the software.
The first month
- Days on site in week one
- 2+
- Operational views set up for your warehouse
- Week 4
- at the latest
- Systems replaced
- 0
- it goes on top
Counted from the day we start on site.
Around 80%
Of the value found early was already inside the operation as it ran, so nothing had to be rebuilt to reach it
From our own work in warehouses, looking at where the early value came from
Let's get you to the quickest value point, then build at the rate your team can absorb. We've learned the hard way that handing a warehouse everything at once overwhelms the people who have to use it. Nor do we chase every last minute of the day, since getting most of it accounted for is enough to manage well and the rest costs more to capture than it tells you.
Intuitive does not equal automatic
The numbers only pay if somebody acts on them, which is why the first months come with our people as well as the software. We work alongside your supervisors and managers in those months, so acting on what Fettle shows becomes part of how the shift runs.
In one warehouse the list of pick faces to change was far longer than a stretched team could get through. Rather than hand over the whole list, we handed it over a day's worth of moves at a time.
Our founder learned the same thing rolling out proof of delivery to 800 drivers, most of them on minimum wage and few of them keen on a new device changing how they worked. It stuck because the way it was introduced got as much attention as the device itself.
What does warehouse labour analytics cost?
Fettle is charged per tracked operator, on the operators actually working, which comes to under one per cent of what an operator costs you. There's also a piece of work in the first months to set it up and embed it with your team. Both figures come out in the demo, once you've seen what it could save.
The options you're probably weighing
Each of these is the right call for somebody, so here's what each one costs you, what it fixes and what it leaves behind.
| Option | What it costs you | What it fixes | What it leaves |
|---|---|---|---|
| Wait for your platform's own AI | Nothing now, then whatever the vendor charges when it arrives | Whatever that vendor ships, on its release timetable | The gap between releases and anything held outside that one system |
| Replace the system | A new licence and the services to put it in, usually the largest bill here, plus the risk that comes with any go-live | A system that has reached the end of its life | The same kind of transactional data in a new format, so the reporting question often comes back |
| Buy a labour management system | A licence, plus engineered standards to build and keep up to date | Measuring people against the standard you set | Whether the process itself should change, which a standard can't tell you |
| Bring in consultants | Day rates for as long as they're with you | The problem they're brought in for, often well | A gap when they go, unless your team can carry the work on |
| Do nothing for now | Nothing up front | Nothing, though it keeps this year's budget free | The gap, which grows as the business asks more of the warehouse |
Fettle goes on top of whatever you run now and carries on if you change it later, so it doesn't have to wait for any of these decisions.
How your data is looked after
Picture Perfect Data, the company behind Fettle, is certified to ISO 27001:2022, the international standard for managing information security. The certificate covers our data analytics, consultancy and software development.
ISO 27001:2022, certificate number 459602024, issued by Citation ISO Certification Limited
See it set up for a warehouse like yours
The demo takes about an hour and is set up for the kind of warehouse you run, with the rate coming out at the end.