B2B market entry in MENA without hiring in every country
B2B market entry in MENA fails on sequencing more than on strategy. Which country first, what has to be local, and the costs that appear after you commit.
Read
A forecast is not a prediction of what you will sell. It is a statement about how well you understand where your deals actually are, and most of them fail at the second part.
Ask a sales team why the quarter missed and you will usually hear that a couple of big deals did not land. Ask what happened to those deals and the answer is almost never that the buyer said no. They went quiet, or legal took longer than anyone allowed for, or the sponsor moved to a new role and the process restarted from a position nobody had budgeted time for.
That distinction matters more than any spreadsheet technique. A forecast that misses because deals died is a qualification problem. A forecast that misses because deals moved is a timing problem, and timing problems are far more common and far more fixable.
Published benchmarks on forecast accuracy are worth knowing but worth handling carefully. A benchmark compilation by Forecastio gathers the commonly cited figures in one place: research attributed to SiriusDecisions that 79 percent of sales organisations miss their forecast by more than 10 percent, and research attributed to CSO Insights that close to 60 percent of forecast B2B deals slip into the following quarter rather than being lost.
Read those as orders of magnitude rather than precise measurements. They are compiled figures rather than primary studies you can inspect, the underlying research is not recent, and vendor-published benchmarks tend to describe the problem the vendor sells against. What they are useful for is the shape of the thing: missing is normal, and the dominant failure is movement rather than loss.
The same compilation puts working accuracy bands at roughly 80 to 95 percent for strong teams, 50 to 70 percent for typical ones, and below 50 percent for teams that are effectively guessing. Before you use those to judge yourself, you need your own number, and most teams have never calculated one.
You cannot improve a number you have not measured, and forecast accuracy is unusually easy to measure because the answer arrives on its own every quarter.
Take the last four closed quarters. For each one, find what the forecast said at the start of the quarter and what actually closed. Express the gap as a percentage of the forecast. That series is your baseline, and its variance matters more than its average, because a team that is 30 percent over one quarter and 30 percent under the next has an average that looks respectable and a process that is useless.
| What you find | What it usually means | Where to look first |
|---|---|---|
| Consistently over-forecast | Optimism at stage entry, or stages defined by rep activity rather than buyer action | Stage definitions |
| Consistently under-forecast | Sandbagging, usually a rational response to how the number is used | How you react to a miss |
| Wildly variable | Too few deals for the average to be stable, or one deal dominating the quarter | Deal concentration |
| Accurate in total, wrong per deal | Errors cancelling out, which will stop working the moment volume drops | Per-deal close dates |
The fourth row is the one that catches people out. A team can hit the number for three quarters while being wrong about every individual deal in it, and that arrangement holds only while the law of large numbers is doing the work for you.
If most misses come from movement, then movement deserves a number of its own. Slippage rate is the share of deals forecast to close in a period that were still open at the end of it.
Track it per stage and the picture usually resolves quickly. Slippage concentrated at one stage is a process problem at that stage, not a rep problem. Procurement and legal are the classic offenders, because they are the two steps that are genuinely outside the seller's control and the two that sellers most often forget to put time in for.
Once you have a slippage rate by stage, you have something better than a gut feel about close dates. If deals at contract stage slip 40 percent of the time by an average of three weeks, that is not pessimism to build into the forecast, it is arithmetic.
Teams reach for a better forecasting model when what is wrong is the data going into it. There is no weighting scheme that rescues stages nobody applies consistently.
None of this works if the underlying records are unreliable, which is why CRM hygiene for small sales teams is a forecasting topic rather than an administrative one. The same goes for the pipeline itself: if it was never built to be measured, see how to build a B2B sales pipeline you can actually forecast.
Pipeline coverage gets used as though it were a forecast, and it is not. It tells you whether there is enough in the pipe to be plausible, which is a different question from what will close.
Coverage of three times target is a rule of thumb that hides an assumption about win rate, and your win rate is knowable. Work out your own coverage number from your own conversion rates rather than inheriting somebody else's multiplier. A team converting at 40 percent needs nothing like the coverage of a team converting at 12 percent.
The most common forecasting error we see is not optimism about whether a deal will close. It is optimism about when, driven by a sales cycle assumption that was never checked against the team's own history.
If deals in your segment take five months and a rep is forecasting one that entered the pipeline six weeks ago, no coaching conversation about that specific deal is as useful as the observation that it is being judged against the wrong clock. Realistic cycle ranges by industry are a starting point, but your own median matters more, and you can calculate it from closed deals in an afternoon.
A forecast becomes more accurate when the data behind it is current, and it stops being useful when producing it takes three days of someone's week. That is a genuine argument for putting the pipeline somewhere it can be read live rather than assembled by hand each Friday. Our colleagues at Rivl wrote a good piece on what "real-time" actually means for a sales dashboard, including the cases where a nightly refresh is genuinely enough and the live version is being bought for the wrong reason.
What tooling does not fix is a stage definition nobody agrees on. Automating a bad process gives you the same forecast faster.
Some of this does not apply to you yet, and it is worth saying which parts.
If you close fewer than roughly 20 deals a year, statistical forecasting is not available to you in any meaningful sense. One deal is a large share of your year, and no method converts that into a reliable number. What you can do is track slippage and be explicit about the range, which is more honest than a single figure carrying false precision.
If your business is genuinely lumpy, with a handful of large deals, forecast the deals individually and name the assumptions rather than aggregating them. And if the forecast is used punitively, expect sandbagging and treat it as a design consequence rather than a character flaw. People report what is safe to report, and that is a management decision before it is a data one.
The realistic goal is not accuracy in the abstract. It is a forecast you would be willing to make a hiring decision on, and knowing how wrong it is likely to be in which direction.
Send us your last four quarters, forecast against actual, and your stage definitions. We will tell you whether you have a slippage problem, a qualification problem or a definitions problem, which are three different fixes that all look identical from the outside.
Book a Free Consultation