Two storefronts in Italy, and nowhere to hide a weak one

Concentration cuts both ways. The good half is that two platforms is a workload a single restaurant can genuinely stay on top of: two menus to maintain, two ad accounts to pace, two sets of ratings to defend. Owners in markets with a longer platform list end up managing shallowly everywhere and properly nowhere.

The other half is exposure. If your Just Eat listing slides down the results page after a bad two weeks of late orders, or your ratings drift far enough that ranking stops favoring you, there is no third channel quietly absorbing the loss. Recovery is slower than most owners expect, because ranking leans on recent performance and the position has to be earned back order by order.

So the first job on an Italian storefront is usually defensive rather than promotional: opening hours that match reality, item availability kept current, prep times set to what the kitchen actually does at Saturday dinner service, and the handful of items generating most of the complaints. Growth spend on top of an unreliable storefront is money spent advertising a bad experience.

No, and copying one storefront into the other is among the most common things we fix. The apps order categories differently, surface modifiers and combos differently, and crop and place photography differently while someone scrolls. An item list that converts on one can read as a wall of text on the other.

Photography is where that gap is widest and where the cheapest wins usually sit. A dish with no image is a line of text competing against pictures, and the platforms' own published benchmarks put real dish photos at up to 20% more clicks and up to 32% more orders. That figure is theirs, not ours — but it is why a shot list for your best sellers comes before almost anything we would ask you to pay for.

The rest of a rebuild is item names and descriptions in Italian, category sequencing so your highest-margin dishes are seen before the scroll dies, and modifier groups that add value instead of confusion. Across the storefronts we manage, menu conversion doubled after a rebuild. It matters most on a high-volume, low-ticket menu, where cheap-to-produce items need combo structure built around them rather than a flat percentage off the top.

With two platforms, the temptation is to leave an offer running permanently on both. It stops being a promotion at that point and becomes your price, and it moves nothing. We plan offers weekly per storefront, margin-checked before they ever reach you, and stagger them so both apps are not discounting the same dish in the same week.

Offer structure matters more than offer depth. Platform-reported benchmarks put buy-one-get-one at roughly 39% more order volume and around 25% more order value, because it grows the basket instead of shaving the ticket. A flat percentage off tends to do the reverse unless it is aimed at a specific item or a specific group of customers.

Every plan is a recommendation with the expected effect attached: which offer, on which items, on which app. You approve or reject it, and nothing reaches your storefront without that step.

OrderLift works remotely from Amsterdam, with manager access to your two back offices. In a two-storefront market that covers more than owners expect: menu structure and copy, photography briefs, category order, opening hours, prep-time settings, availability, the offer calendar, the hour-by-hour ad budget and the weekly reporting all live in software you can watch us work in.

What it does not cover is your Saturday service. Ratings and prep-time accuracy are earned in the kitchen, and in a market this concentrated they are worth defending harder than any promotion. We can tell you which dishes travel badly and which items drive the complaints, and hold budget back until the storefront converts, but we cannot cook.

Our 60+ partner restaurants are in the Netherlands, not in Italy, and every result we claim for ourselves comes from them. The dish-photo and buy-one-get-one figures on this page are the platforms' own published benchmarks, and we label them that way.