Customer Concentration Risk: When One Client Can Sink a Company

In late 2013, a company called GT Advanced Technologies announced a deal to supply Apple with sapphire material, widely expected to end up in iPhone screens. The stock ran hard on the news. Less than a year later, GTAT filed for Chapter 11, and the bankruptcy documents showed why the deal was never the win investors assumed. GTAT had taken on the buildout costs and the production risk, while Apple, by then its overwhelmingly dominant customer, had committed to almost nothing. Apple was under no obligation to buy the sapphire at all.

That collapse is the extreme version of customer concentration risk, which shows up in milder forms all over the market. A company that depends on one or two customers for a large share of revenue is carrying a risk that stays invisible while things are good. Concentrated revenue can even look like a strength. The relationship is stable, the orders are predictable, the sales team knows the account inside out. Then the customer leaves, or gets acquired, or vertically integrates, or just starts squeezing harder, and the economics of the whole business shift at once.

The useful part is that this risk gets disclosed. Companies are required to tell you about it, in specific places, in fairly standard language. You just need to know where to look and what to do with what you find.

Where the Disclosures Live

Under ASC 280, the segment reporting standard, a company has to disclose when any single customer accounts for 10% or more of total revenue, along with the amount. The standard doesn't require naming the customer, which is why so many filings talk about Customer A and Customer B. Separately, SEC rules for the business description push companies to name a customer when the relationship is material enough that losing it would seriously hurt, so you will sometimes get the name in Item 1 of the 10-K even when the footnotes stay anonymous.

In practice, check three places in the 10-K:

• The segment or revenue footnote. This is where the 10% disclosure usually lives, often under a heading like concentration of revenue or major customers.
• The concentration of credit risk footnote. This one covers receivables. A customer that is 20% of revenue but 45% of accounts receivable is also a collection risk, and receivables from a dominant customer stretching out over time is an early warning on its own.
• Item 1A, the risk factors. Look for phrases like a limited number of customers or the loss of any major customer. Most of it is boilerplate, but the specifics wrapped around the boilerplate, like named customers or contract end dates, are worth pulling out.

When a filing only says Customer A, you can usually work out who it is. EDGAR full-text search lets you search every filing for a company name, which shows you which suppliers mention it. Press releases about contract wins, trade press, and earnings call transcripts fill in the rest. Analysts on the call will often just ask.

How Much Concentration Is Too Much

Rough rules of thumb I use when reading a filing:

• No customer above 10%. A diversified base. Losing any single account stings but doesn't change the story.
• One customer at 10 to 20%. A real dependency, but survivable. If the account walks, the company has a painful year and then recovers.
• One customer at 20 to 40%. Material risk. Losing the account means restructuring, and the customer knows it, which weakens the company's hand in every negotiation between them.
• One customer above 40%. Effectively a single-customer business, whatever the total customer count says. Survival depends on keeping one relationship healthy.

The percentage alone undersells the picture, though. Who the customer is matters as much as how big it is. A 30% customer that is the US government behaves very differently from a 30% customer that is a struggling retailer or a cash-burning startup. So ask two questions about the customer itself: how healthy is it, and how easily could it replace this supplier? A dominant customer with cheap alternatives is far more dangerous than one that would need years and real money to switch away.

What a Dominant Customer Does to Margins

Most concentrated relationships never end in a GTAT-style collapse. The more common damage is slow, and it comes from bargaining power. A customer that knows you can't afford to lose them will use that knowledge, politely and relentlessly.

It shows up in three ways. First, price. Dominant customers ask for concessions at every renewal, and suppliers with concentrated exposure have very little room to say no. Anyone who has sold into Walmart or supplied Apple knows the annual cost-down conversation. Second, payment terms. The customer stretches from 30 days to 60 to 90, the supplier finances the gap, and working capital quietly turns into a loan the supplier is making to its biggest account. Third, scope. Extra services, faster turnaround, custom work, all absorbed without a matching price increase, because nobody wants to reopen the pricing conversation.

Each of these looks manageable on its own. Compounded over years, they can leave the flagship account barely profitable, and management doesn't always know it, because customer-level cost-to-serve accounting is rarer than it should be. When the largest customer keeps growing as a share of revenue while gross margin drifts down year after year, this squeeze is the first thing I would check for.

What the Contract Actually Guarantees

Companies like to point at long-term agreements as proof that a concentrated relationship is safe. Sometimes that holds up. A five-year contract with minimum purchase commitments puts a supplier in a genuinely different risk position than a relationship running on purchase orders the customer can simply stop issuing.

The details decide which one you are looking at, so dig for them in the filings and on the calls:

• Minimum volumes or take-or-pay commitments. Without minimums, a supply agreement is often just an option the customer holds.
• Termination for convenience. If the customer can exit without cause, the stated contract length is close to meaningless as protection.
• Exclusivity. If the supplier is barred from serving anyone else, the risk concentrates further while the upside stays capped.
• Change of control clauses. If the customer gets acquired, does the agreement survive the acquirer?

This is exactly where GTAT investors got hurt. The headline said multi-year Apple deal. The actual terms, which only became public in bankruptcy court, let Apple walk away while GTAT carried the capital costs and exclusivity obligations. When a company declines to describe even the basic structure of its most important contract, that silence tells you something too.

Concentration Looks Different by Industry

Some industries run concentrated by nature. Defense primes sell mostly to the US government. Auto parts suppliers live off a handful of automakers. Component makers often ride a single big design win, and Cirrus Logic has disclosed for years that Apple accounts for the large majority of its sales. In these industries the market already knows, prices the risk to some degree, and the right benchmark is industry peers rather than the broader market.

The setups that deserve extra attention are the ones where concentration is unusual for the industry, or unusually extreme within it. Plenty of consumer products companies name Walmart as a 10%-plus customer, and that's normal for the category. A consumer products company doing, say, 30% of its revenue through a single retailer is in a different position, because it is carrying a bet on one buyer's shelf space and goodwill that its peers are not carrying. Outliers within a peer group are where this analysis earns its keep.

The Trend Tells You More Than the Snapshot

One year's disclosure tells you where concentration stands. Five years of disclosures tell you where it's heading, and the direction usually matters more than the level. Pulling the major customer footnote from the last five 10-Ks takes maybe ten minutes on EDGAR, and laying the numbers side by side is one of the highest-value steps in this whole exercise.

Say the largest customer was 15% of revenue five years ago and sits at 30% now. Growth has been driven by deepening dependence on one account, and every incremental dollar makes the eventual renegotiation or loss more painful. Now run it the other way. A company that has taken its top customer from 30% down to 18% while still growing total revenue has been doing the slow, unglamorous work of diversification, and its risk profile is improving even though the headline number still looks elevated.

Where new revenue comes from is the same signal in different clothes. If most of the recent growth came from the existing top customer, the dependency is compounding. If it came from new customers, the company is buying itself options.

What Losing the Customer Does to the P&L

A surface reading gets this part wrong. Losing a 30% customer costs far more than 30% of profit, because the fixed costs stay behind.

Say a company reports $200 million in revenue, with its largest customer at 30%, or $60 million. Gross margin is 40%, so gross profit is $80 million. Operating expenses run $60 million, which leaves $20 million of operating income, a healthy 10% margin. Now remove the customer. Revenue falls to $140 million, and gross profit at the same 40% falls to $56 million. But the $60 million of operating expenses is mostly salaries, rent, and systems that don't shrink on the same timeline, so the company is suddenly running a $4 million operating loss. A 30% revenue event wiped out more than all of the operating income.

So when a company has a customer above 20%, run the scenario explicitly. What happens to revenue, margins, and cash flow if the customer winds down over twelve months? Can the company service its debt through the transition, and does anything in the covenants trip? If the answer looks like financial distress, today's valuation should carry a discount for that tail risk, and the discount should scale with how hard the lost volume would be to replace. This is a big part of why two companies with identical growth and margins can reasonably trade at very different multiples.

What Serious Management Teams Do About It

Concentration is partly a management quality question, and the disclosures let you grade it. Look for evidence that the company is spending real money on diversification while the anchor relationship is still healthy: sales capacity pointed at new accounts, new geographies or channels, products that pull in a different customer segment, and integration work that raises the big customer's cost of switching away. A management team that talks about its largest customer on every call but never reports progress on new ones is showing you where the energy actually goes.

Timing is the point I would underline. Diversification is cheap to start when the anchor relationship is strong and cash is coming in, and brutally expensive once the customer begins to wobble. GTAT never really had the option, since its agreements restricted selling sapphire to anyone else. Most concentrated companies do have the option, for years, and either use it or spend that time hoping the next renewal goes well.

A workable routine for the next 10-K you open: search the document for 10% and concentration, read the major customer and credit risk footnotes, note whether customers are named, pull the same disclosure from the four prior years to get the trend, scan the risk factors and the last couple of earnings calls for contract terms, and run the loss scenario on anything above 20%. The whole thing takes about half an hour per company, and it covers a risk that most screens and models skip entirely.