Nobody knows how many market entries fail. Here’s what the evidence does say.

Most market entries don’t fail. They stop. Our last piece looked at four decades of trade data and found that new export flows usually end within a year or two, quietly and early. This piece asks the harder question: what separates the entries that last from the ones that don’t?

Ask ten advisers why international expansion fails and you will get ten confident answers. Many will come with a statistic attached. “70% of international expansions fail.” “Half of cross-border deals destroy value.” “Four in ten companies fail when they go abroad.”

We went looking for where those numbers come from, and then for what the stronger evidence actually shows. It is drawn from peer-reviewed research, World Bank and NBER datasets and decades of studies on subsidiaries and exporters. The short version: there is no reliable universal failure rate. But there is a consistent pattern in who survives, and most of it is visible before the money is committed.

First, the statistics

The most quoted number, that 70% of international expansions fail, has no traceable origin. It appears on consultancy blogs and in conference slides, usually without a citation. One site that uses it concedes the figures are “directional” and that there is no single way to measure them. We could not find a study that defines “expansion”, defines “failure” and counts both.

The numbers that do hold up are narrower, and each measures something different:

  • Exporters. The World Bank’s Exporter Dynamics Database covers 45 countries. The median share of new exporters still exporting a year later was 43% in 2006–08, ranging from 23% to 61% by country (Cebeci et al., 2012). In Argentina, 47% of firms that started exporting in 2006 lasted only one year (Albornoz et al., 2012).
  • Cross-border acquisitions. KPMG’s 1999 study of 107 companies behind large cross-border deals found 53% destroyed shareholder value within a year, measured by share price against their sector. That is the likely source of “half of cross-border deals fail”. It is one sample, one period and one definition (KPMG, 1999).
  • International retail. Studies of foreign retailers entering the UK since 1850 find that 73–75% eventually left. But that exit rate covers more than a century, and leaving after 60 years is not the same as failing (Burt, Dawson & Sparks, 2008).

None of these is a failure rate for companies entering new markets. An exporter that stops after a year may have run a sensible test and stepped back. An acquisition can lose value on the stock market and still be the right strategic move.

The honest finding is that nobody knows the universal failure rate, because the question is badly defined. What the evidence does show is more useful than a headline number. Early exits are common, survival after the first year or two is high, and the survivors do very well. Colombian customs data shows only about a third of new exporters survive their first year. After that, annual survival rises to 80–90%, and within a decade the survivors account for almost half of total export growth (Eaton, Eslava, Kugler & Tybout, 2008).

So the useful question is not “what are the odds?” It is “what changes the odds?”

Distance still decides more than market size

The most consistent finding in international economics is that distance cuts trade, and it has not gone away. A meta-analysis of 159 studies puts the typical distance effect close to proportional: double the distance and trade roughly halves. A shared language adds around 60%, and a colonial link more than doubles trade (Head & Mayer, 2014). An earlier review of more than 1,400 estimates found the distance penalty “has remained persistently high”, internet or no internet (Disdier & Head, 2008).

Pankaj Ghemawat’s CAGE framework makes the point commercially: distance is Cultural, Administrative, Geographic and Economic, and a market ranked by GDP alone looks far more attractive than it is (Ghemawat, 2001). For goods where buyers need to trust the supplier, and not just compare prices, these barriers are higher still (Rauch, 1999).

But close is not the same as easy. In a study of 32 Canadian retailers that entered the US, the “obvious” market next door, only seven operated successfully. The researchers called it the psychic distance paradox: “assumptions of similarity can prevent executives from learning about critical differences” (O’Grady & Lane, 1996).

For UK companies that is a direct warning about the US, Ireland and Australia. A shared language removes one barrier and hides the rest: procurement rules, buying committees, channel structures and price expectations. The markets that feel familiar are the ones where businesses most often skip the homework.

The foreignness penalty is real but it fades. Foreign trading rooms in 47 countries were less likely to survive than local rivals, and the penalty changed over time (Zaheer & Mosakowski, 1997). New foreign-owned firms in Portugal were no more likely to fail than domestic ones once size, strategy and industry were accounted for (Mata & Portugal, 2002). Being foreign is a handicap you can manage. It is not a verdict.

Experience helps, but only the right kind

Prior international experience improves survival. But generic experience (“we already sell abroad”) counts for much less than experience in the same or similar markets.

A study of 225 foreign entries by Dutch firms found that cultural distance shortened how long ventures lasted. Previous experience in the same country, or in culturally similar ones, lengthened it (Barkema, Bell & Pennings, 1996). Across 3,080 subsidiaries of Japanese firms, host-country experience improved survival, but on its own it did not guarantee profit (Delios & Beamish, 2001). Experience keeps you alive long enough to learn. It does not do the learning for you.

The sequencing evidence is some of the most practically useful research in the field:

  • Most exporters start with one market. In Argentina, 79% of new exporters began in a single destination. Those that survived their first year grew sales there by 190% the following year, then added markets. The authors’ explanation: firms don’t know how well they will do abroad until they try, and success in one market predicts success in others (Albornoz et al., 2012).
  • The second market should look like the first. Chilean exporters found that entering a country similar to a previous one, in geography, language or income, cut the cost of entry by 69–90% (Morales, Sheu & Zahler, 2019).
  • Speed has a cost. Across 741 foreign expansions by Dutch firms over 25 years, a faster pace, an irregular rhythm and a wider geographic spread each reduced the profit expansion delivered. Organisations can only absorb so much new learning at once (Vermeulen & Barkema, 2002).

What about companies that go global from day one? The evidence says high variance. Among roughly 5,800 Belgian SMEs, global start-ups committed the most to each market and grew fastest, and also had the highest failure rate (Sleuwaegen & Onkelinx, 2014). In a UK sample, firms that internationalised at founding had lower survival overall, but the gap disappeared once their competitive strategy was accounted for (Mudambi & Zahra, 2007). Going wide early is not wrong in itself. Going wide without a clear reason to win in each market is.

Entry mode matters less than fit, and entry is a search problem

There is a long debate about whether to acquire, build your own subsidiary, form a joint venture or use a distributor. The raw data seems to favour building your own: acquisitions and joint ventures entering the US exited more often than greenfield subsidiaries (Li, 1995). But when Myles Shaver controlled for the fact that firms choose the mode that suits their own strengths, the survival advantage “disappears” (Shaver, 1998). There is no universally best entry mode. There is only a mode that fits what you already do well.

Intermediaries are a rational way in, particularly for hard markets. In Chinese export data, distributors and trading companies handled a larger share of trade to markets that were more distant, smaller or more document-heavy. Many firms used them first and then went direct (Ahn, Khandelwal & Wei, 2011). The risk, which David Arnold documented across roughly 250 market entries, is a familiar cycle: the distributor delivers early sales, growth stalls, and the manufacturer eventually takes back control at a cost (Arnold, 2000). A distributor is a route to market, not a strategy for one. It is worth deciding at the outset what would trigger taking it back.

The deeper finding is about what entry actually is. Recent research on exporters suggests it is mostly a search for buyers, and building buyer relationships takes years. French firms exported almost only to markets where they already had a contact, and used those contacts to find the next ones (Chaney, 2014). For Colombian exporters to the US, losing their accumulated customer base and market knowledge was estimated to cost around $3.4m, about a third of their total export value (Eaton et al., 2021). The revised Uppsala model reaches the same conclusion from management research: the core barrier is not being foreign but being an outsider to the networks where business gets done (Johanson & Vahlne, 2009).

That has direct consequences for resourcing. Entry costs are largely sunk: in Colombian plants, having exported the year before raised the likelihood of exporting by up to 60 percentage points (Roberts & Tybout, 1997). A market budgeted as a twelve-month launch, when the evidence says relationships take several years to build, is under-resourced before it starts. Kogut’s research on joint ventures points to the better model: treat early commitments as options. Commit small, and scale on real signals rather than plans (Kogut, 1991).

Government support helps at the margin. A study of export promotion agencies in 103 countries found they raise exports, especially where information gaps are large. Returns fall sharply as budgets grow, and the median return in OECD countries was far lower than in developing economies (Lederman, Olarreaga & Payton, 2010). For UK firms, trade missions and export support are best used for introductions, not as a substitute for a market case.

Localise the price before you localise the brochure

The research on standardising versus adapting your proposition is less decisive than either camp would like. A recent review calls the evidence “mixed and inconclusive” (Mandler et al., 2021). Standardisation pays when demand is similar across countries and scale economies matter (Schmid & Kotulla, 2011). But one element stands out. In a meta-analysis of 23 studies, adapting price had the strongest link to performance of any part of the marketing mix (Brei et al., 2011).

The pricing evidence explains why. Exporters routinely charge different prices for the same product in different places. Portuguese exporters charged more in richer and more distant markets, even for the same product from the same firm (Bastos & Silva, 2010). A study of identical products sold online by four global retailers found that price gaps between countries were mostly set when the product launched in each one, and then persisted (Cavallo, Neiman & Rigobon, 2014).

That second finding is the one worth remembering. The launch price in a new market is sticky. A price calculated from home-market list price plus freight and duty, rather than tested against what local buyers will pay, can stay wrong for years. We found no rigorous study of how often SMEs get this wrong. But the direction of the evidence is clear: willingness to pay differs by market, the best-performing firms price to it, and the first price you set tends to stick.

The best-known retail failures also look different up close. Walmart left Germany in 2006 with about a $1bn loss, having failed to beat Aldi and Lidl on price (CBS News). Target Canada closed 133 stores in 2015 with a $5.4bn pre-tax loss, largely because of supply-chain failures (Reuters). Tesco took a £1.2bn charge to exit its US Fresh & Easy chain in 2013 after extensive pre-entry research (The Grocer). They are usually told as stories about culture. They were at least as much about price position and operations: things that were either testable before entry, or only discoverable once in it.

What you can know before entry, and what you can only learn by entering

Taken together, the evidence splits the uncertainty in market entry into two kinds.

The first kind is knowable in advance, cheaply and with reasonable confidence. Distance, in all four CAGE dimensions. Whether the target customer exists there in enough numbers. What they use today and what it would take for them to switch. What they will actually pay. Who signs off, and how procurement works. What regulation and certification stand in the way. The psychic distance paradox, the pricing research and the retail failures all point at problems in this category: things that could have been found out, and weren’t.

The second kind is only discoverable by entering. The Argentine exporters did not know whether they would succeed until they tried; that is the whole point of the sequential exporting research. Relationships with buyers and partners, how the organisation copes with operating at a distance, and how competitors respond cannot be researched from a desk. They are learned, and the learning costs money.

QuestionKnowable before entry?How it is answered
How far is this market, culturally, administratively, geographically and economically?YesDesk research, trade data, CAGE assessment
Does the target customer exist here, in enough numbers?YesSegment data, customer interviews
What would make them switch from what they use now?LargelyBuyer interviews, competitor mapping
What will they pay, and who signs?LargelyWillingness-to-pay testing, buying-process mapping
Is the route in open? Regulation, certification, channel marginsYesRegulatory review, channel economics
Can we build the buyer relationships we need, and how fast?PartlyOnly proven through early, limited entry
Will a partner or distributor perform?PartlyTested through a bounded trial with clear triggers
Can our organisation execute at a distance?NoLearned in market; stage the commitment

This is where the argument of our earlier piece and the research meet. If market-entry risk is front-loaded, and most of it sits in the first year or two, then the value of pre-entry work is in shrinking the first category to near zero, so that the money spent in market is spent on the second category alone.

Validation is not about eliminating uncertainty. That is impossible, and pretending otherwise is how businesses talk themselves into overconfident launches. It is about deciding which uncertainties are cheap enough to resolve before you commit, and which are worth paying to discover once you’re in.

Before you spend serious money: what you should actually know

If the evidence were boiled down to the questions a leadership team should be able to answer before committing to a new market, they would look like this:

  1. How far away is it really? Not in miles, but on all four CAGE dimensions. And if it feels close, what are we assuming is the same that might not be?
  2. Does it resemble a market where we already win? The strongest predictor of success in a new market is similarity to one you have already succeeded in.
  3. Who exactly is the customer, and what would make them switch? Interest is not demand. Name the trigger, and the person who owns the risk of changing supplier.
  4. What will they pay? Tested locally, not converted from the home price list. The launch price will be hard to move later.
  5. How will we find and keep buyers? Entry is a search problem. Whose network gets us in, and what happens to those relationships if a partner leaves?
  6. Does our entry mode fit our strengths? And if we use a distributor, what will trigger us taking back control?
  7. Have we funded the years, not the launch? Most exits happen early. The budget has to carry the business through the window where the evidence hasn’t yet arrived.
  8. What would make this a no? Set the revenue, margin and payback thresholds before looking at the evidence, not after.

None of these guarantees success. Nothing does. But the research is consistent on one point: the companies that survive their first years abroad tend to be the ones that entered markets they understood, at a pace they could absorb, with commitments they could scale up or step back from. Most of that is decided before the first contract is signed.

If a new market is on your agenda in the next twelve months, I’m always happy to talk it through. — Louisa Seymour, LS Advisory

References

Failure and survival rates

Distance and market selection

Experience, sequencing and pace

Entry mode, commitment and networks

Localisation and pricing

Leave a Reply

Your email address will not be published. Required fields are marked *