In 2006 Blake Mycoskie started selling canvas shoes out of his Venice Beach apartment with a promise attached: buy a pair and a child somewhere gets a pair. TOMS hit $1 million in sales in it’s first year and $100 million within five. Business schools wrote it up. Warby Parker copied the structure, Bombas copied it, Skechers copied it hard enough to get sued over it. For about a decade, if you read anything about purpose-driven branding, you read about TOMS.
By 2019 the company was carrying so much debt that creditors took control from Bain Capital and from Mycoskie himself, after years of falling sales. In April 2021, on it’s fifteenth anniversary, TOMS formally ended one-for-one. It now commits at least a third of annual net profits to grassroots organizations working on mental health, access to opportunity and gun violence.
I bring this up because the one-for-one model still shows up as the flagship example in branding advice written this year, usually phrased as something you should consider imitating. The company that invented it stopped doing it and the reasons it stopped are more useful than the model ever was.
What actually went wrong, since the lessons are specific
Three separate things broke and they map onto three different pieces of standard branding advice.
The differentiator stopped differentiating. Once every competitor had a giving model, the giving model stopped being a reason to choose TOMS. Meanwhile the product underneath, essentially one canvas slip-on, had not moved much. A brand position built on a mechanism anyone can copy has a shelf life measured in whatever time it takes for someone to copy it.
The operations got heavier than the story. TOMS distributed more than 95 million pairs of shoes over the program’s life and their chief strategy and impact officer later described the giving network in plain terms, telling Glossy that <cite index=”56-6″>”We had hundreds of shoe giving partners and that’s too much to manage”</cite>. Every marketing promise has an operations bill attached and the simpler the promise sounds to a customer, the larger that bill tends to be.
The impact claim did not survive scrutiny. Critics argued for years that donated shoes addressed a symptom rather than poverty itself and that free goods undercut local vendors. Here’s what I find genuinely admirable, though: in 2010 TOMS commissioned outside academic researchers to study exactly that and did not anonymize itself in the results. The findings showed a real but modest effect, roughly one lost sale for a local vendor per twenty pairs donated. Small. Not zero. Most companies in that position never commission the study and the ones that do usually insist on anonymity.
That third point is the one worth sitting with. TOMS got the mechanism wrong and the transparency right and the transparency is why we can discuss the mechanism at all.
So what survives from the standard playbook
Here’s the uncomfortable part for anyone selling branding strategy: most of the advice is fine. It is also so widely known that following it produces no advantage. Understanding your segment, having a coherent story, standing for something. All true, all worth doing, all currently being done by your competitors. Table stakes do not win tables.
The values thing needs a durability test
If you’re going to build a brand position on values, the question is not whether customers like the value. It’s whether you can still afford the commitment when sales fall 30% and whether it stays yours once three competitors announce the same thing.
TOMS failed both tests. The new model, giving a share of profits rather than a unit of product, passes at least the first one, because a percentage of profit scales down when profit does. That’s not cynicism, it’s the difference between a commitment that survives a bad year and one that becomes a liability during it.
Personalization is mostly a data problem wearing a marketing costume
The advice to personalize is old enough now to be furniture and the reason most brands do it badly has nothing to do with strategy. It’s that personalization requires data infrastructure, clean records and someone accountable for the recommendations being good rather than merely present. Recommendation engines that suggest the thing you already bought are worse than no personalization, because they demonstrate the brand is watching without paying attention.
I’d rather see a company do one personal thing well, remembering a size, honoring a preference, than deploy a system that generates the appearance of attention.
The digital advice ages fastest of all
Every version of this article names whatever technology is current at the time of writing and calls adoption essential. Chatbots. AR try-on. Whatever comes next. Some of these matter enormously in specific categories, virtual try-on genuinely reduces returns in eyewear and apparel and mattered much less in others. The tell of a weak strategy is naming the technology first and the problem second.
What I don’t know
Whether TOMS’s new model actually produces more good than the old one is not something I can verify from the outside and I notice most coverage takes the company’s word for it. Giving a third of profits to grassroots organizations is harder to audit than counting shoes, which is both the strength of the model and the reason to hold judgment. It may be a real improvement. It may also be a system with fewer countable failures.
The useful takeaway isn’t a strategy. It’s a habit: when you read a branding case study, check what the company is doing now. A remarkable share of the examples that circulate as best practice are describing a phase the original company already exited, sometimes at considerable cost. Advice that never gets rechecked stops being advice and becomes folklore.