“The Process That Got You to £40m Will Break at £150m”: Inside Itsu Grocery’s Scaling Reality

We caught up recently with Sam Howell, Supply Chain Director at Itsu grocery to talk about the operational realities of scaling a fast-growth consumer brand. From supplier constraints and S&OP redesign to AI scepticism and changing talent needs, the conversation explored the uncomfortable but necessary shift from entrepreneurial agility to operational maturity.

Ed Lawson, The Retail Hive: Your business is scaling at an extraordinary pace. What have been the biggest wake-up calls in realising that the systems and processes which worked at £40m won’t work at £200m?

Sam: The reality is that the process that works at £40m probably isn’t fit for £80m, and definitely won’t be fit for £150m. You’re constantly having to adapt.

But equally, you can’t just leapfrog overnight into some perfect enterprise model. You have to evolve iteratively, which means the goalposts are always moving internally.

A huge part of that becomes finding the right people. You need employees who are comfortable with ambiguity, flexible enough to evolve with the business and willing to keep changing how they work as the company matures.

That affects the whole organisation, but supply chain probably feels it more intensely than most functions. Marketing maybe experiences it even more because consumer behaviour changes so quickly, but supply chain definitely gets hit hard during scaling.

Ed: What does that feel like within the business? 

Sam: I think the biggest shift has been moving from an unconstrained world into a constrained world. Historically we’d send forecasts to suppliers and assume things would broadly work themselves out. Now we’re having to ask much harder questions, like can suppliers actually support us in six months? And what does future capacity really look like?

Making sure we’re operationally fit for future growth has become one of the top priorities in the business.

Supplier capacity becomes critical very quickly. Logistics should theoretically be easier to scale, but even there we’ve had issues. This year we oversold pallet capacity quite significantly and the problem with underforecasting is that it can take months to recover operationally.

Inbound capacity has become a real pressure point too, particularly with frozen products. If you don’t forecast accurately where the growth will come from, suddenly processes like container tipping start becoming bottlenecks.

Ed: You’re redesigning your S&OP process and moving forecasting ownership away from sales and into demand planning. That’s often as much a cultural shift as an operational one. What’s been hardest?

Sam: Culture change is always the hardest part, people get comfortable with the way things have always worked and naturally nobody loves change. But as the business matures, the processes have to mature with it.

We’re moving from having two demand planners this year to four demand planners plus a demand planning manager. The goal is to take some responsibility away from sales and improve forecast quality overall.

The key for me has been making sure the sales director is fully engaged in the process and ensuring the wider SLT supports the investment behind it. For this kind of change to work, both sides have to see the value. Sales teams need to understand why the shift benefits them and leadership needs to be willing to invest behind it properly.

Ed: A lot of scaling businesses dream about demand outstripping supply, but supplier constraints can become a problem very quickly. How are supplier relationships evolving as the business grows?

Sam: We’ve worked really hard to make relationships much more personal and deeper.

We still absolutely have transactional suppliers, and that’s fine. We understand we’re not a major percentage of every supplier’s business and equally some suppliers aren’t strategically critical to us.

What matters is identifying the handful of suppliers where deeper collaboration genuinely matters, particularly where there are capacity constraints.

For those suppliers, we’re investing much more time face-to-face. We’ve visited suppliers across Japan, China and Germany because those relationships become incredibly important as volume scales.

Some of our suppliers in China have met with us five times already this year between procurement and supply chain teams.

Ed: Tell us what those conversations look like?

Sam: We’re now having conversations around longer-term contracts, guaranteed volumes and suppliers investing in machinery, automation, technical capability and account management specifically to support our growth.

But obviously that only happens if we also commit more strongly to them.

If you’ve gone from shipping ten containers a month to fifty, one account manager probably can’t support that anymore. The relationship has to mature operationally on both sides.

Ed: There’s huge noise around AI in forecasting and supply chain right now. Where are you seeing genuine value versus industry hype?

Sam: I absolutely see the value in AI, but I also think the industry is getting ahead of itself in places. At the moment, with the quality and consistency of data most businesses actually have, we’re only comfortable using AI in relatively narrow use cases where we’re confident it’ll genuinely add value.

I still haven’t seen an affordable AI forecasting engine that convincingly delivers transformational forecasting improvement. Where I see more immediate value is around transparency of data, decision-making tools and surfacing insights faster rather than fully autonomous forecasting.

That said, the skillsets inside the team are definitely changing, when we recruit into demand planning now, we’re increasingly looking for people with stronger data analytics capability rather than purely traditional forecasting backgrounds.

Longer term, the goal isn’t necessarily to keep increasing headcount. It’s to gradually evolve the role itself so planners spend less time manually analysing spreadsheets and more time making commercial decisions.

Ed: One of the useful things about peer conversations is realising even very large businesses are often wrestling with similar issues. Are there any industry trends currently making you stop and think?

Sam: Honestly, one surprise for me has been how many very large businesses still don’t have particularly joined-up data strategies.

You assume huge companies have perfect visibility and beautifully connected systems, but a lot are still operating with ten different systems that barely talk to each other.

There’s also definitely a shared scepticism around AI in some areas. Most people can see the potential, but a lot of businesses are still figuring out where the real practical value sits.

And talent is a huge conversation everywhere. Particularly around how younger workers are entering the industry with very different skillsets and expectations compared to even ten years ago.