The percentage you asked for (retail-specific, UK, and defensible)
There is no official UK statistic that says “AI will kill X% of retail businesses”. If anyone claims a precise number, they’re guessing with confidence.
That said, if you’re asking for a realistic range (not a fairy tale), here’s the best I can justify from UK business churn + what AI is doing to retail economics:
- Over the next 10 years: ~6% to 18% of UK retail businesses could plausibly go out of business with AI-driven competitive pressure as a major contributor (not the only cause, but a meaningful one).
That’s on top of “normal” failure. UK business death rates (all sectors) run close to ~1 in 10 per year in recent ONS demography data (though “death” is not identical to insolvency and includes dormant/closures).
Why this isn’t a neat number (and why retail is extra exposed)
Retail doesn’t die from “AI” — it dies from margin collapse
A retail business rarely fails because a robot walked in and took the keys. It fails because:
- footfall drops
- online customer acquisition becomes unaffordable
- price-matching becomes instant and brutal
- returns/fraud rise
- bigger players squeeze suppliers and delivery costs
- your shop becomes a showroom for someone else’s algorithm
Retail is a spreadsheet industry wearing a cardigan. Anything that improves pricing, targeting, logistics, and conversion will favour the biggest operators fastest.
UK retail is already under strain before AI finishes warming up
- UK government insolvency stats show 23,938 registered company insolvencies in 2025 (England & Wales).
- Industry commentary on insolvencies stresses levels remain elevated versus pre-pandemic “norms”.
- And retail employment has been shrinking: BRC cites 2.88m retail jobs in Dec 2024, down 249,000 over five years, with tech/automation part of the story.
In short: plenty of retailers are already running on fumes. AI simply makes the race faster.
What “AI taking over the marketplace” means in UK retail
1) Discovery gets monopolised
Search results, shopping feeds, marketplace rankings, social algorithms… increasingly AI-curated.
If you’re not winning inside those systems, you effectively don’t exist. This is where smaller retailers get quietly suffocated: not beaten on product, beaten on distribution.
UK competition regulators have explicitly raised concerns that AI markets could concentrate power among a small number of incumbents (the same firms that already control major digital channels).
2) Price competition becomes ruthless (and instant)
AI-driven dynamic pricing and competitor monitoring means:
- your niche price premium gets tested every minute
- your promos get copied quickly
- your slow-moving stock gets exposed faster
Independents don’t lose because they’re “bad”. They lose because they can’t play the same pricing game at the same speed.
3) Operating cost advantage widens (big firms get even bigger)
AI makes the strongest retailers stronger by improving:
- demand forecasting
- stock allocation
- shrink/fraud detection
- customer service automation
- marketing creative at scale
BRC survey reporting (via Reuters) shows retailers planning to boost productivity and invest in automation as labour costs bite.
4) Customer expectations rise, but willingness to pay doesn’t
Customers get used to:
- instant answers
- instant comparisons
- fast delivery
- frictionless returns
They do not get used to paying you more for it.
So why 6%–18%? (A transparent, retail-specific logic)
The UK already has heavy business churn
ONS business demography shows a UK-wide business death rate of 9.8% in 2024 (all sectors).
That’s not “insolvencies”, but it’s a clear indicator of how much turnover the economy already has.
Retail has more “AI-replaceable” economics than most sectors
Retail has lots of functions that AI can compress:
- basic product copy/content
- merchandising decisions
- customer service
- paid ads optimisation
- conversion-rate optimisation
- procurement support
- forecasting and replenishment
So the slice of closures where AI is a major contributor is likely higher in retail than the economy average.
My split (10-year view)
- Low end (6%): AI mostly boosts efficiency; weak retailers still fail, but AI isn’t the main reason.
- Mid case (10%–12%): AI accelerates online/marketplace consolidation; independents get priced out of attention.
- High end (18%): distribution becomes even more pay-to-play; marketplaces and large chains fully weaponise automation, and a lot of “small but generic” retail disappears.
Who is most at risk in UK retail (and why)
Highest risk: “I sell the same stuff as everyone else” retailers
- Gift shops with identical supplier catalogues
- Generic fashion/accessory resellers
- Commodity phone accessory/electronics add-ons
- Dropship-like e-commerce with no brand moat
- Any store relying on Google/Facebook ads but with thin margins
AI makes it easier for competitors to:
- copy your listings
- outrank your ads
- undercut your pricing
- identify your best-selling SKUs and replicate them
High risk: independent non-food high street without a destination hook
If you’re selling standard products with no genuine reason to visit:
- you become a showroom
- footfall declines
- rent/rates don’t
Medium risk: supermarkets and big-box (they won’t die, but they will hollow out jobs)
Big retailers can deploy automation. The pain shifts to:
- head office roles (planning, reporting, admin)
- store labour hours
- entry-level jobs
Recent reporting tied BRC survey findings to retailers planning cuts and automation investment due to rising costs.
Lower risk: “retail with a human reason to exist”
- specialist fitting/repair/alterations
- local service wrapped around product
- community-led destination retail
- true expertise categories (where advice matters and returns are expensive)
AI helps these businesses too — but it doesn’t replace the core value as easily.
Expert quotes you can use
On the direction of travel: more cost pressure, more automation
- Reuters reporting on a BRC survey highlighted that retailers are planning productivity drives and automation investment as labour costs become a top concern.
On market power risks in AI (why smaller retailers may get squeezed)
- The CMA’s foundation model work flags concerns that AI could reinforce the power of large incumbent firms.
What I’d do if I were an English retailer trying not to be roadkill
The survival checklist
- Stop selling “generic”: either niche down hard, or build a brand people search for by name.
- Own your customer list: email/SMS/loyalty. Algorithms change; first-party relationships endure.
- Use AI to cut the boring costs first: product content, FAQs, internal reporting, simple customer queries.
- Compete on trust + service: clear returns, fast responses, honest stock information.
- Treat marketplaces like a tax: useful for volume, dangerous for dependence.
References (UK / primary where possible)
Business churn / closures
- ONS Business demography (business “death rate” 2024):
- UK Government company insolvency statistics (2025 annual):
Retail labour + structural pressure
- BRC: retail job losses over five years (ONS-based):
- Reuters/BRC survey on labour costs and automation intentions:
AI concentration / market power
- CMA foundation models update paper + review pages:
