The difference between site-level and group-level GP — and why both matter
A group GP figure tells you how the portfolio is performing overall. A site-level GP figure tells you which sites are carrying the others. You need both.
HOPS Team
Product & Operations
Group GP and site-level GP are two different answers to two different questions.
Group GP answers: how is the portfolio performing overall? It is the number that goes in board presentations, investor reports, and year-end accounts. It reflects the combined trading of all venues and smooths out the variation between sites into a single measure.
Site-level GP answers: how is this venue performing? It reflects the specific trading conditions, menu mix, team, and cost structure of one operation. It is the number that a GM manages against and that a regional manager uses to compare performance between sites.
Both are necessary. The problem arises when operators have only one of them.
What gets hidden in the group figure
A group trading at 65% blended GP might contain a site at 71%, a site at 67%, and a site at 55%. The strong performers are carrying the weak one. The board presentation looks acceptable. The site that is structurally under-performing continues to do so without visible pressure, because the group number never falls low enough to require a conversation.
This is not hypothetical. It is the standard operating condition for multi-site groups that do not track site-level GP separately and consistently. The group figure becomes a comfort that obscures the range of performance within it. Understanding how to consolidate reporting across multiple sites is the structural fix that makes the site breakdown possible in the first place.
The inverse problem is equally common: a group whose overall GP is under pressure that cannot be located because site-level visibility does not exist. Something is wrong somewhere. The group number is telling you that. The conversation about where and why cannot happen without the site breakdown.
What site-level GP requires
A site-level GP figure is only meaningful if it is produced from the same inputs and the same methodology as every other site in the group.
If site A calculates GP from a weekly stock take and site B calculates it from monthly invoice totals, the two figures are not comparable. Site A's GP reflects actual consumption. Site B's reflects purchasing activity, which diverges from consumption whenever stock levels change. Comparing them tells you something, but not what you think it tells you.
The same applies to category structures. If food and beverage are separated at site A but blended at site B, the blended GP at site B is a different number from the blended GP at site A. They might look similar and mean something completely different.
Consistent methodology across every site is the prerequisite for site-level comparison to be meaningful. This is not a finance technicality: it is the difference between management information and misleading information.
The role of site-level GP in performance conversations
When site-level GP is visible, clear, and produced consistently, the management conversation changes.
A regional manager reviewing three sites with GPs of 69%, 64%, and 61% has a starting point for three different conversations. The 69% site is performing well: what is it doing that the others are not? The 64% site is close to target: what is moving it, and is it trending in the right direction? The 61% site is below range: is it a cost problem, a revenue mix problem, a team problem, or something specific to that site's trading conditions?
None of those conversations can happen without the data. And the quality of the conversation depends on the reliability of the data, which depends on the consistency of how it was produced.
Benchmarking across the estate
Site-level GP is also the foundation for benchmarking: understanding what good performance looks like for a specific type of operation within your group, and using that as a target for sites that are below it. This is also where separating food and beverage GP within each site adds a further layer of precision, because kitchen performance and bar performance can differ significantly between venues even when the headline site GP looks similar.
A group with five casual dining sites that all have broadly similar menus, service models, and cost structures can use the top-performing site's GP as an informed target for the others. Not a generic industry benchmark, but a real figure from a comparable operation in the same group, achieved under similar conditions.
This is more powerful than any external benchmark. It is specific to your concept, your supplier relationships, your team calibre, and your cost base. If site A can do it, sites B, C, and D can understand why they are not.
“You can really tell HOPS has been built by operators. They understand our needs and provide a solution that is exactly what we want, instead of us having to adapt and change for a system.”
Dominique Fernandes
Head of Operations, Mildreds
Where Hops fits
Hops produces site-level GP from connected inventory, purchasing, and sales data at each site, using a consistent methodology applied across the group. The group view aggregates those site figures. The management hierarchy means each person in the structure sees the level relevant to them: site GP for GMs, cluster views for regional managers, group consolidated for the FD. For operators who want to understand what the full group-level P&L looks like when it is built on consistent site data, that is the next layer of financial visibility to build.
The site-level and group-level figures are the same number viewed at different granularity. They are not produced by different processes that then have to be reconciled. That consistency is what makes the comparison meaningful.
Frequently asked questions
What is the difference between site-level GP and group-level GP?
Site-level GP reflects the trading performance of a single venue, accounting for its specific menu mix, team, and cost structure. Group-level GP combines all sites into a single figure that shows overall portfolio performance. Both are needed because the group figure can look acceptable while hiding a site that is significantly under-performing.
Why does my group GP look fine but feel wrong?
A group GP figure is an average, and averages hide variation. A strong site running at 71% can carry a weak site at 55% to produce a blended group figure that looks reasonable. Without the site breakdown, the under-performing site never receives the scrutiny it needs and the gap rarely closes on its own.
How do I make site-level GP comparable across my restaurant group?
The figures are only comparable if they are produced by the same methodology: the same category structures, the same counting frequency, and the same treatment of items like deposits and voids. A site calculating GP from weekly stock takes is not directly comparable to one calculating it from monthly invoice totals, even if the numbers happen to look similar.
How should I use site GP figures in performance reviews?
Site GP gives a regional manager a concrete starting point for three different conversations: investigating what the top performer is doing well, understanding what is moving the middle performer, and identifying whether the lowest performer has a cost, revenue mix, or operational problem. Hops produces consistent site-level GP across all venues so those conversations are grounded in comparable data — see how at hopshq.com.
Can I benchmark one site against another in my group?
Yes, provided the figures are produced consistently. A top-performing site in the same group is a more meaningful benchmark than any industry average, because it reflects your specific concept, supplier relationships, and cost base. The comparison is only valid when the methodology is identical across sites.
Tags
Built for operators
See how operators are actually using Hops.
We could tell you what Hops does. Instead, read what the people running their businesses on it have to say.