Diversification

A company that runs one business is at its most efficient while that business is going well. People and budget point the same way, and there is little to argue about over what comes first. The trouble arrives when the one thing wobbles. If the budget of the main customer segment is cut in half in a single year, company revenue is cut in half with it — not because the product got worse or a competitor won, but because one event outside the company transfers straight onto the income statement.

That setup usually assembles itself quietly. Adding people to whatever is working is the correct call every time it is made, so three years in, 70 percent of revenue comes from one product or one segment. While revenue keeps climbing this doesn't look like exposure. It is only on the day someone in a board meeting first asks what happens if that segment cuts its budget next year that it becomes clear nobody prepared an answer.

Diversification is the decision to prepare that answer in advance. What separates it from most growth decisions is its purpose: it exists so the company stays standing when one part collapses, not so the company earns more. Which is why diversified companies generally post worse numbers than focused ones during good years. Those lowered numbers are the price of surviving the bad ones.

Diversification means finding what current revenue is riding on and building a second axis that doesn't depend on the same thing. There are four steps.

1. Count where revenue concentrates. Break it down by product, by segment, and by region, and one or two cells are usually overwhelming. Any cell above 60 percent is, by itself, the company's entire exposure. The count is worth doing more than one way, because a company selling to three large customers in the same industry looks spread out by customer and is not spread out at all.
2. Choose a direction to expand in. Sell a new product to the same customers, sell the same product to a new segment or region, or go new on both. The further along that list, the larger the upside and the higher the chance of failure. Selling a second product to people you already talk to can reach its first contract in a few weeks; selling a new product to a new segment means finding out who the buyer even is, so plan on a year.
3. Check how much overlaps with what already exists. If the sales organization, the technology, and the brand carry over, it's related diversification; if the only shared resource is cash, it's unrelated. A document tool company selling workflow automation is the former. The same company going into property rental is the latter.
4. Actually split the resources. This is where the bill comes due. Resource allocation is the work of deciding how much headcount and budget goes where, and giving the new axis 10 percent means taking 10 percent off the existing one.

Step 4 is where most attempts come apart. Budget is a line item, so money moving elsewhere leaves a mark; people get quietly reassigned the moment something urgent happens. Someone whose name sits on the new business can spend four days a week on incidents in the old product and the books will show nothing wrong. Diversification usually fails not because the direction was wrong but because nobody actually went there.

The word surfaces in three kinds of conversations.

1. Calling out revenue concentration. When the weight of one segment comes up in a board or investor discussion, diversification is the direction offered as the answer.
2. Deciding whether to start something new. New business development is the work of building a business in a different area; diversification is the reason given for doing it. The same proposal gets asked for one set of numbers when growth is the argument and a different set when risk is.
3. Weighing the opposite. A focus strategy pours everything into one narrow area, and for a company of around 50 people that is usually the right answer. Both words come out in the argument over when there is finally enough slack to widen, and the honest version of that argument is about headcount rather than ambition.

"Seventy percent of revenue comes from one segment. It's time to talk about diversification."

Pointing out a structure where one wobble takes the whole company with it.

"That's not diversification, that's losing focus. We can't carry two products with the headcount we have."

Agreeing with the direction while arguing there is nothing left to split.

"It's related, so the sales org carries over. The only new team is on the engineering side."

Explaining that the overlap with existing resources is wide enough to keep the cost down.

At Pagedeck the word came up at last quarter's board meeting. Sofia had pulled the contract picture together, and more than half of annual contract value was coming from startup customers. The problem was that those customers are the first to cut tool budgets when the funding market cools.

Grace asked for three options to be worked up: selling the same document tool to larger enterprise accounts, selling a new workflow automation product to today's customers, and widening nothing at all while going deeper on the current product. Everyone agreed that 45 people cannot do two of them at once, and the decision moved to next quarter. What the room actually argued about was not which direction was bigger but how many people could be pulled away.

The most common reading is that diversification reduces risk, full stop. The picture of several baskets comes to mind easily enough. But what drops is the risk of one market turning, and what appears in its place is the risk of failing at something the company has never done. Until the second axis stands on its own, both risks are carried at once. Leave that stretch out of the plan and a move meant to reduce risk produces the weakest 18 months the company will have.

Treating feature expansion as business expansion is the other frequent mistake. Adding tables to a document tool is selling the same product better to the same customers; it does not create a second axis. The test is what remains when the work fails. A feature that misses leaves existing revenue untouched, while an axis that misses takes the headcount and the months with it. File both in the same drawer and what gets reported to the board as diversification is really a feature list, and the concentration question that started it all is still sitting there a year later.