I had lunch at Dunelm Mill last week. It was average. Though it sounds like a restaurant, Dunelm Mill sells homewares, curtains, duvets, pillows, mirrors, small items of furniture and preposterously large been bags. The cafe is a way of getting customers to linger.

Dunelm Mill’s wares are also average, not particularly flash, but not tat. OK, a carved wooden ornament spelling ‘love’ is tat, but mostly Dunelm Mill’s shelves are stacked with objects you could easily imagine in most homes, at prices most people could easily imagine paying. The company’s motto is "Simply value for money".

Walking past Venetian blind displays, and packs of ten plastic hangers for ''2.99, it seemed Dunelm has positioned itself in a retail sweetspot. Its wares are big enough and personal enough that you’d probably want to go into a store to choose them but not so big and expensive, that you wouldn’t in a recession. The bed and sofa shops next door are empty.

With goods piled high and massive signs leading customers to them, it feels like a John Lewis warehouse with wholesale prices attached to the products.

There’s plenty of evidence for the retail sweetspot theory in the results. We’ve had two recessions, yet Dunelm has grown strongly. Sales have risen 70% since it opened its 80th superstore in 2007, and profits have more than doubled. It now operates 124 stores, 115 of them out of town superstores like my local. In 2012 Dunelm earned a 30% return on tangible assets.

It’s more than able to finance its own growth. The company had no debt on June 30, and is proposing to return 32.5p a share, pretty much the whole of this year’s profit figure, to shareholders in November 2012.

Investors must weigh the tremendous profit potential in a successful retail format rolling out nationally against the tendency for growth stories to come to a stuttering halt once the nation is saturated with stores.

Dunelm is obviously thinking about this. ‘Catchment modelling’ suggests that saturation point is 200 superstores, and it’s already experimenting with smaller formats. But the company admits that as it gets closer to the 200 store limit, it will be harder to make money from each new superstore. Whereas new stores opened in the last three years have paid for themselves within 31 months, the company says in future payback might take 48 months, allowing for the fact that new stores will take sales from existing ones.

Management’s frankness is another indicator the company is in good hands, but maybe Dunelm is approaching a dangerous, or at least less profitable, phase for shareholders. On a PE of 18 the company must grow to justify its share price and although Dunelm probably will for some time an end is, possibly, in sight.

Some retailers find ways to keep on going, though. In addition to its fashion shops, for example, which it runs very efficiently, Next is rolling out a new format of homeware stores. They’re flashier than Dunelm’s and there’s one just round the corner from the Dunelm Mill on the edge of Cambridge.