Every insurer wants SME premium and almost none earn it. The gap is not a product problem or a demand problem — it is a distribution-cost problem, and that is fixable.
Ask any commercial insurer where they want to grow and the answer is SME. Ask them where their loss ratios are healthiest and the answer is often the same. Then look at how many small businesses actually hold adequate cover, and the numbers collapse. The SME segment is the industry's hardest easy market: attractive on every actuarial slide, stubbornly unreachable in practice.
The short version of this piece: the gap is not caused by bad products or absent demand. It is caused by distribution economics. A policy that costs the same to sell as a corporate programme but earns a fraction of the premium will always lose the fight for an insurer's sales capacity — until the cost of selling it approaches zero. That is a channel problem, and channel problems have engineering solutions.
The paradox in one paragraph
SME risks are individually small, reasonably standard, and diverse enough to pool well. A portfolio of shop, clinic and contractor policies is close to an underwriter's ideal book. But each policy might carry a few thousand riyals of premium, and the traditional way to sell it — an agent visit, a proposal form, a back-and-forth on sums insured — costs nearly as much as selling a policy fifty times the size. The economics push every distribution channel upmarket, and small businesses are left with a choice between buying nothing and buying something generic that nobody explained.
Why small businesses do not buy
Talk to owners rather than insurers and the gap looks different. Three patterns come up repeatedly.
- Time poverty. The owner is the finance department, the HR department and often the delivery driver. An insurance purchase that demands a meeting and a form loses to literally everything else on the desk.
- Illegibility. Commercial policy wordings are written for brokers. An owner comparing two business packages usually cannot tell what differs, so they default to price — or to postponement.
- Misplaced confidence. Most small firms have never had a serious loss, and human beings price risk from memory. The fire, the flood, the lawsuit are abstractions until they are not.
None of these is irrationality. They are rational responses to a buying process that was designed for someone else.
Why insurers do not sell
The supply side has its own honest reasons. Acquisition cost per policy barely changes with premium size, so commission structures push intermediaries toward larger risks. Underwriting an SME properly requires information — turnover, stock values, safety measures — that is expensive to collect by questionnaire and often wrong when collected. And servicing thousands of small accounts through channels built for hundreds of large ones erodes whatever margin the pricing promised.
So both sides behave sensibly and the market stays broken. In Saudi Arabia the stakes of that failure are rising: Monsha'at counted roughly 1.2 million SMEs in the Kingdom in its Q1 2023 monitor, and Vision 2030 policy aims to lift the sector's contribution to GDP from around a fifth to 35 percent. A growing SME economy with thin insurance penetration is accumulating uninsured risk at national scale.
The SME gap is not a demand problem wearing a disguise. It is a cost-of-sale problem, and cost-of-sale problems yield to software.
Distribution is the missing piece
The reason we call this an easy market wearing hard clothes is that the blockers are all in the channel, not the risk. Consider what changes when insurance attaches to software a business already uses.
The data collection problem disappears: an accounting platform already knows turnover; an HR system already knows headcount; a point-of-sale system already knows the trade. The time-poverty problem shrinks: an offer surfaced inside an existing workflow, pre-filled and priced, asks minutes rather than meetings. And the acquisition-cost problem inverts: the platform already owns the relationship, so the marginal cost of presenting a relevant offer approaches the cost of an API call.
This is the same embedded logic that moved motor and travel cover into checkouts, applied to commercial lines. It is earlier in its development — commercial products are less standardised, and sums insured need more care than a flight date does — but the direction is the same.
The honest caveats
Embedding does not dissolve every difficulty. Complex risks still need advice; a factory with unusual processes should talk to a broker, and no checkout flow should pretend otherwise. Pre-filled data reduces errors but does not eliminate the underinsurance problem — a business that grew 40 percent since its accounts were filed needs its sums insured to grow too. And an offer inside a workflow can be ignored just like an email can; relevance and timing decide attach rates, not the mere existence of a button.
The claim is narrower and stronger: for the standard covers most small businesses need first — property, liability, medical for staff — the main reason they go unsold is that selling them costs too much. Put the offer where the business already works, price it from data that already exists, and the hardest easy market starts behaving like what it always was underneath: a large, diverse, insurable population that nobody could previously afford to reach.