How to get recommendations across to management

An auditor wrote nine recommendations, and every one was sound. They explained them to the director for 40 minutes, the director said "good, we'll look at them", and a month later nothing had changed. The director was not against them; they simply did not fully grasp why they were needed, what they cost, who would do them and when they would pay off. Writing a good recommendation is not enough: it must be delivered in a form management can accept and set in motion.

One recommendation, one line

A recommendation answers four questions: what is done, who does it, by when and what result it brings. A poor version: "Improve order in the warehouse." A good one: "Introduce a second approval for stock adjustments. Owner: chief accountant. Deadline: 30 days. Result: the route for a hidden shortage is closed." The first cannot be carried out; the second can be checked.

Cost and benefit: with figures

Management measures a recommendation in money. Suppose the auditor recommends barcode scanners for picking. Cost: 3 scanners at 2.5 million so'm each, plus 2 million so'm for setup, 9.5 million so'm once. Benefit: picking errors cause 12 orders a month to come back, each costing about 300,000 so'm (re-dispatch, courier, a discount to the customer), which is 3.6 million so'm a month. Payback: 9.5 / 3.6, about 3 months. A calculation like this sounds far more convincing than "we need scanners".

Options, and the recommended one

Where possible, give two or three options and pick one:

1. Do nothing: no cost, but the loss of 3.6 million so'm a month continues.
2. Three scanners: 9.5 million so'm, payback about 3 months.
3. Mobile terminals and software configuration: noticeably dearer, a fuller result.

"We recommend the second, because it is enough for the size of the warehouse, and the third can follow later." Management keeps its choice, but the road is shown.

Preparing for the presentation

Before the meeting, do three things: agree with the audited party, prepare answers to management's questions and start with a quick win.

Discuss it first

Management should not hear the report for the first time in a meeting. First meet the head of the audited area: are the facts right, is the recommendation workable, what obstacles are there? Their view goes into the report as a separate line. The recommendation is then received not as an attack but as a result of cooperation. Thirty minutes is enough for the meeting with management: 5 minutes of conclusion, 15 of discussion and 10 for a decision.

Prepare answers to management's questions

Four questions usually come up: why now? What happens if we do nothing? Who will spend how much time on it? How will it affect our other work? Write the answers down before the presentation. The last matters most: if a recommendation takes two days out of an owner's working week, say so openly, otherwise it will stall with "no time".

Start with a quick win

There is no need to start everything at once. Put a cheap, quick recommendation first (for example, adding a second approval or renewing labels): the first success builds faith in the harder steps. Split costly recommendations into stages and move on to the next once the result of the previous one is seen.

The presentation itself

The presentation rests on two things: language management understands and a clear request for a decision.

Speak management's language

In the auditor's language: "There are weaknesses in the control procedures." In management's: "We lose 3.6 million so'm a month, and 9.5 million so'm will stop it." Translate qualitative description into so'm, time and risk. Drop the jargon; the three most important recommendations are enough for page one.

Ask for a decision

End the presentation with a question: "Do you approve these three recommendations, with the owners and deadlines shown?" The answer can be yes, no or conditional, but not a vague "we'll see". Once a decision is made, enter it in the tracking table at once and set the date of the next update: from then on the recommendation is work, not a document.

Common mistakes

• Twenty recommendations of equal significance.
• Language of blame: "there is no second approval in the process" instead of "the storekeeper isn't doing it".
• Hiding or understating costs, so that trust is lost later.
• No owner or deadline named.

Conclusion

A recommendation works not when it is written in a report, but when management has said "yes" and an owner is appointed. OmborAudit hands over a recommendations plan with priority, cost, owners and deadlines and is ready to discuss it with management.