StarTech Multimedia and Supplies
Month end went from three weeks to four days
| Measured | Before | After |
|---|---|---|
| Month end | 3 weeks | 4 days |
| Stock variance | 11% | 1.4% |
| Quotations below margin | unknown | 0 |
| Re-typed figures a month | about 400 | 0 |
Before
- Quotations in a spreadsheet one person understood
- Stock counted twice a year and wrong both times
- Payroll re-typed from a second spreadsheet
- The accountant found the errors, six weeks later
After
- Every quotation built from the specification, with the margin checked before it leaves
- Stock reconciled weekly, with unaccounted differences stopping rather than being adjusted
- Payroll and the ledger reading the same figures
- Month end closed in four days
We built it because nothing we could buy fitted. We sell it because the firm down the road has the same problem.
How we measure what changed
Two firms is two firms. Here is what that is and is not evidence of.
They are two results. One of these firms wrote the software, which is a conflict of interest and is stated at the top of their entry rather than left for you to work out.
We did not measure them. We asked, they counted, and where they could not produce a before we have left the row out rather than guessing at it.
Both took longer to get running than either expected, and in both cases the reason was data that had to be cleaned before it could be loaded. We will say more about that on a call than a page like this can.
What changed, for each of them
Two firms is a small number and the band below says so. It is still worth reading them against each other rather than one after the other: a print and electrical supplier of thirty-four people and a training provider of twelve have almost nothing in common, and the last column says the same thing twice. Work that was being re-entered stopped being re-entered, and figures checked once a year started being checked weekly.
| Firm | Trade and size | What they started with | Where it is now |
|---|---|---|---|
| StarTech Multimedia and Supplies | Print, ICT and electrical · 34 people | Quotations in a spreadsheet one person understood, stock counted twice a year and wrong both times, payroll re-typed from a second spreadsheet | Every quotation built from the specification with the margin checked before it leaves, stock reconciled weekly, month end closed in four days |
| A training provider in the Central Region | Education · 12 people | Admissions on paper in a box, registers collected at the end of the week when they were collected at all, results assembled by hand into a spreadsheet | Applications arriving into one list with the gaps visible, registers taken on a phone offline in the classroom, results locked until every mark is in |
The two right-hand columns are the firms' own words, in the order they gave them. The second firm asked to stay unnamed and is drawn here as they asked to be.
Five things to check on any supplier's case study, ours included
You will read several of these before you decide, and every one of them is written to persuade you. These are the five checks that separate a measurement from an advertisement — apply them to this page first, where the conflict of interest is ours.
- Look for the base under the percentage
A forty per cent improvement with no starting number is a sentence, not a figure. Every change on this page is written as a before and an after in the same unit, because that is the only form you can argue with.
- Check who did the counting
We did not measure these firms. We asked, they counted, and where they could not produce a before we left the row out. A supplier who measured their own customer's improvement has written a different kind of document.
- Look for the conflict of interest
One of the two firms here wrote this software. That is a real conflict and it is stated at the top of their entry rather than left for you to notice. If a case study has one and does not say so, it is the most important thing missing.
- Look for what went wrong
Every implementation has a bad fortnight. Both of these took longer to get running than either expected, for the same reason, and it is written down. A case study with no setbacks in it is a press release.
- Ask for a customer they did not choose
The firms on this page agreed to be here. The useful call is with somebody in your own trade and your own size, chosen by you from a list rather than offered — ask for that, from every supplier.
What happens when you ask to speak to one of them
The call is arranged by us and then it is yours. We are not on it, there is no script, and the firm is told only which business is asking and what they do.
Who you will be speaking to
Whoever ran the change inside that firm — usually the person who owns the figures rather than the person who signed the contract. They are the one who can tell you what the fortnight after go-live felt like.
How long it takes to arrange
A few working days, because we are asking somebody for a favour rather than booking a resource. If the firm closest to your trade is unavailable we will say so rather than substituting a different one quietly.
Whether we are on the call
No. We make the introduction and then step out. A reference call with the supplier listening is a call where the awkward question goes unasked, which makes it worthless to you and flattering to us.
What they have agreed to discuss
Anything. Neither firm has been asked to stay off a subject, including what it cost them, what went wrong and what they would do differently. If a question gets a careful answer, that is their judgement and not our instruction.
What we are told afterwards
That it happened. Not what was said. If you want to tell us something that came out of it, that is yours to offer.
Measure your own first year and we will publish it
This page grows one firm at a time and every firm on it had to agree to be counted. The offer is the same for anybody reading: pick the figures that matter to your business before you start, and we will help you record them so the comparison afterwards is real.
Three or four, no more, and each one countable by somebody in the business without a report being built for it. Days to close a month. Stock differences per count. Quotations that went out below cost. Applications that arrived incomplete.
The before is the half everybody forgets, and it cannot be reconstructed afterwards. An afternoon with the old records, written down and dated, is the whole cost of being able to prove something later.
One period is the period everybody was watching. Two is where the number settles into what it will be.
The figures and the name are two separate permissions and you give them separately, afterwards, having seen exactly what would be published.
Being measured is optional and being named is a second, separate decision. One of the two firms here chose to be measured and to stay unnamed, and that is drawn exactly as they asked.
We will put you on a call with a firm your size
Not a reference we chose for the occasion -- a firm in the same business as you, who will tell you what the first three months were like. Ask us and we will arrange it.