There is a middle path that gets overlooked in this conversation. The platform and the display hardware do not have to come from the same company. The Howard Company is Square's chosen drive-thru partner [verify approved wording; delete this note before publish], and it builds systems that integrate with the major restaurant platforms rather than requiring one, which is why its digital display line is built to be technology agnostic. Buying hardware you own outright and software you can change is a different risk profile than buying both from one vendor on one contract.
Takeaway: free hardware is a financing structure, not a price. Get it priced both ways, and ask what the hardware is worth on its own, before you compare.
"What do we spend on signage" means four different things depending on who is asking. Mixing them up is the most common reason a signage budget comes in wrong. A facilities manager wants a unit price. A development lead wants a per store package. A CFO wants a fleet number. Those are different figures, and only the first one is easy to find.
Takeaway: if you are benchmarking a fleet number against a component price, you will be wrong by an order of magnitude. Start by deciding which level you are budgeting.
Because the number depends on the building more than the brand, and because no two disclosures draw the line in the same place, as The Howard Company found reading the filings for this benchmark.
Franchise disclosure documents are the most detailed public source on restaurant startup costs, and even there the signage line means something different in every filing.
Two figures we read directly in the filings themselves:
Those two numbers are 12 times apart and neither one is wrong. One buys every sign on a commercial parcel, interior and exterior, digital menu boards and installation included. The other buys one exterior sign package on a landlord's fascia, within a shopping center sign band and a municipal size cap. They are not two prices for the same product. They are the sign scope of a leased bay against the sign scope of an owned site.
The pattern holds across the filings we could read. Inline and leasehold formats disclose exterior signage in the low thousands to roughly $40,000. Freestanding pad sites with drive-thrus disclose $23,000 to $300,000. Within each group, brands bundle differently: Jersey Mike's breaks interior branding into its own line, and Sonic keeps digital menu board housings out of its signage line and puts them in a technology line instead.
Public company filings do not close the gap either. No major restaurant company breaks out signage, menu boards, or reimage spend as a capital expenditure category in its annual report, so there is no top down number to check a bottom up estimate against.
Takeaway: when you compare your signage budget to a peer's, confirm you are both counting the same things, and confirm you are comparing the same building. Ask whether the pylon, the drive-thru boards, the interior graphics, the content software, the integration work, and the site work are inside or outside the number.
Site work is the item most often left out. Trenching, conduit, concrete footings, bollards and a primary power drop are construction line items, not signage line items. They are usually quoted by a general contractor or an electrical sub, they land in a different budget owned by a different person, and a signage figure that excludes them is not comparable to one that includes them. Installation labor sits in the same gap. Before you benchmark anything, settle where the line between the sign and the site is drawn.
Across five fiscal years, the median signage project ran between $4,800 and $11,800 depending on how many locations the buyer operates, and the largest brands wrote the smallest orders.
Across five fiscal years, the median signage project ran between $4,800 and $11,800 depending on how many locations the buyer operates, and the largest brands wrote the smallest orders.
Source: The Howard Company invoiced order records, fiscal years 2022 through 2026 year to date, drawn from 6,141 projects across all bands. One manufacturer's book of business, not a market census. Bands are defined by the number of locations a buyer operates.
The counterintuitive result is at the bottom of the table. Operators with 1,000 or more locations show the lowest median project value in the set, roughly half what a three location operator writes for a single job.
That is not a volume discount. It is a difference in how the buying is organized. A national brand rolls out across hundreds of sites over months or years, and each purchase order covers one site or one phase of the rollout, so any single order is small even when the program behind it is very large. A three location operator remodeling one restaurant orders the whole package at once, so that single order carries everything.
Note the spread in the middle bands. Operators with 76 to 300 locations show a median near $9,100 and a 90th percentile near $35,900. Any single average for this category will mislead you. Budget with a range.
These are purchase order benchmarks, not per store rollout costs. A 100 location operator does not spend 100 times the median. It also does not spend the median once. Multiplying in either direction gets you a number you cannot defend, and it is the most common way a benchmark like this gets misused.
What the same records show at the program level, without the dollar figures, is the shape of the thing:
The median customer year contains one project in every band. Most buying years are a single job at a single site.
Active programs place several times that. Operators in the 76 to 300 and 301 to 1,000 bands commonly place four to twelve separate orders in a single buying year.
So annual program spend runs well above the single project median, and a rollout year runs well above that again.
So the project median tells you roughly what one site's worth of work costs. Fleet spend is a different calculation, and the budget section below gives you the arithmetic for it.
For per unit component pricing, see the indoor menu board cost breakdown and the digital drive-thru cost breakdown. To price your own project, try the drive-thru price calculator.
Operators did not stop investing in customer facing visual systems. They stopped building new structures for them. That distinction matters for a 2027 plan, because it changes what your signage dollars are buying.
Ground up construction got harder. Retail space under construction was 64.2 million square feet in the first quarter of 2026, against a ten year average near 90 million, with construction costs up 15 to 20 percent, per Franchise Times, August 2026. Brands shifted toward inline, endcap, second generation and drive-thru only formats.
Unit counts still grew. Technomic Top 500 domestic unit count rose 1.8 percent in 2023, 1.6 percent in 2024, and 1.4 percent in 2025. Growth slowed. It did not reverse.
Capital moved to remodels. Wendy's reduced its Build to Suit program by about $20 million in 2025 and redirected the funds to technology and marketing. Burger King committed roughly $2 billion to remodels while its US unit count declined, targeting 85 to 90 percent modern image by 2028. Church's Texas Chicken put its remodel program at about $400,000 per store.
Remodel mandates set the clock. Subway extended its remodel deadline from 7 years to 10 and set a $100,000 minimum. Burger King's threshold is $650,000 per remodel. Jack in the Box added a cosmetic tier at $10,000 to $20,000 per location.
Every one of those programs names signage and menu boards as a component. Not one of them discloses what share of the remodel the signage is. That is the single most useful number in this category, and it does not exist publicly.
In The Howard Company's own order book, the pattern shows up as more projects, each smaller: more separate orders year over year, at a lower median value each. Digital displays have also overtaken outdoor structures as the larger share of the work, a crossover that happened in 2024. This is one manufacturer's experience and should not be read as an industry wide measurement, but it is consistent with a market reallocating from new builds toward refresh work.
Between the content software, the monitoring, the warranty, the content labor, the integration work and the replacement cycle, the recurring cost of a digital display is one of the least documented numbers in this category, and in The Howard Company's experience it is the line most often missing from a capital plan. A static board is a purchase. A digital board is a purchase plus a subscription plus a job somebody has to do.
Sources: The Howard Company published pricing,digital drive-thru cost and indoor menu board cost. Static kit figure from the Popeyes Louisiana Kitchen 2025 franchise disclosure document.
The static comparison in the last row is worth sitting with. A static menu board program is not free to run either. Popeyes discloses $200 to $300 per month for static kits, one of the very few published recurring signage costs anywhere. And the print itself is only part of it. Every menu change means design time, whether that is your own team or an agency on the clock, then vendor coordination, then a production lead time, then freight to every location, and then somebody at each store has to take the old panels out and put the new ones in during business hours. The reprint itself is the cheapest part of a reprint.
Digital displays need a stable network connection, and where that connection comes from is a decision worth making on purpose. In practice most displays run on the site's existing wired network, and the cost lands inside a network bill nobody attributes to signage. That is fine right up until it is not.
The question to settle early is segmentation. Put the displays on the same network segment as the point of sale and you have pulled them into the scope of your payment card compliance work, which is a bigger conversation than a menu board deserves. Put them on the guest network and you have made your menu dependent on guest wifi. The normal answer is a separate segment or VLAN for the displays, which costs configuration time rather than a monthly fee, with cellular kept as a backup path for sites where wired service is unreliable.
None of that is expensive. All of it is easier before installation than after. Ask your integrator what the displays will connect through, and ask your own network team whether that answer changes your compliance scope, before either team assumes the other has handled it.
Either you pay someone hourly to do it, or you pay your own staff to do it. There is no version where it costs nothing. Both paths are legitimate and the right answer depends on how often your menu changes and whether you have design capacity in house.
Outsourced. Predictable, professional, and it happens whether or not your team has a busy week. Published rates run $100 to $150 per hour, with a larger block of hours for the initial menu build. Easy to budget, and easy to under-scope if you only plan for the initial build and not the ongoing changes.
In house. Cheaper per change on paper, and faster when the person is good and available. Two risks, and neither one is cost. The first is attention: a regional marketing coordinator with eleven other priorities will get the price change done and skip the daypart optimization, the promotional rotation and the seasonal refresh. The second is turnover. When the one person who actually knows the software leaves, the capability leaves with them, and the menu quietly stops changing until someone notices. If you run it in house, train a second person and write the process down.
The failure mode is the same either way. Digital signs that never change are just expensive signs. The whole return on a digital system comes from what the static one could not do: changing by daypart, testing promotions, pulling an item the kitchen ran out of. If nobody owns that work, you have bought a more expensive way to display a fixed menu.
The Howard Company offers content management as a service and also supports customers who run it themselves. Maintenance and professional services did not exist as revenue lines in our order book before fiscal 2024 and are now a steady share of it. The market is starting to buy this as a service rather than assume it away, which is a reasonable signal that it was being underestimated before.
Takeaway: price the content work before you buy the screens, and name the person or the vendor who owns it. A digital program without a content owner underdelivers regardless of how good the hardware is.
This is the category most consistently missing from signage budgets, and it is the one that can hold up a launch. A digital menu board that does not know your prices is a slideshow. Making it know your prices means connecting it to your point of sale, and depending on which point of sale and which content management platform you run, that connection ranges from a configuration setting to a development project.
Four integrations come up repeatedly, and each has its own cost shape:
Point of sale to menu board. So that a price change made once appears everywhere, and so that items your kitchen has run out of come off the board automatically. Whether this is turnkey or custom depends entirely on the specific pairing of your point of sale and your content platform. Some combinations have a supported connector. Some require development work on one side or the other, and someone has to pay for it.
Order confirmation. The screen at the drive-thru that shows the guest what was heard. It reads from the order stream, which means it has its own connection to the point of sale, separate from the menu board.
Voice AI ordering. If you are deploying it, it has to sit between the guest, the order confirmation screen and the point of sale. Every one of those handoffs is an integration.
Pricing and daypart data. Zone or market level pricing, and menus that change by time of day, mean the board has to pull from whatever system holds the truth about prices, which is not always the point of sale.
The cost shows up in three places, and only the first one usually gets quoted:
Setup and configuration, per location or per program. The Howard Company's published pricing puts point of sale integration setup for an order confirmation system at starting around $1,500.
Development, when the pairing you have chosen does not already have a supported connector. This is a one time project cost with a real range, and it is worth getting quoted before you commit to the pairing rather than after.
Ongoing. Every time your point of sale is upgraded or your menu structure changes, the integration needs revalidating. Budget it as maintenance, not as a one time build.
Two questions to ask before you sign anything. First: is there a supported, already built connector between this specific point of sale and this specific content platform, or is somebody writing code? Ask for the names of other operators running that exact pairing in production. Second: who owns the integration when it breaks, the point of sale vendor, the content platform, or the signage manufacturer? Get that in writing while you still have leverage, because a menu board showing yesterday's prices is a store level problem within hours.
Your signage line stops appearing in the capital plan at all. It does not get cheaper. It moves into a subscription and a payment processing rate, where it is much harder to see, compare, or exit.
Several restaurant technology platforms run their hardware businesses at or below breakeven and recover the cost through recurring software and payments revenue. This is disclosed in their own filings:
Toast reported hardware and professional services revenue of $180 million against $400 million of cost of revenue in fiscal 2025, a gross loss of $220 million, a figure the company states in its own earnings release. The line was negative in 2024 and 2023 as well. On its first quarter 2026 call, Toast's chief financial officer described that gross profit as negative 13 percent of recurring gross profit streams while "leaning into our customer acquisition momentum."
PAR Technology earned roughly 23 percent gross margin on hardware in fiscal 2025. This is not how every platform operates, and the difference is worth knowing before you assume a bundled quote reflects what the hardware is worth.
This economic model has now reached the drive-thru. Toast acquired outdoor menu board manufacturer Delphi Display Systems in February 2023, and in April 2026 launched a drive-thru package built on that hardware, including 55 inch high brightness outdoor screens and an order confirmation display. Per Nation's Restaurant News, customers must run Toast's point of sale to use it.
None of this makes bundling the wrong choice. It makes it a choice that belongs in front of a chief financial officer rather than inside a technology renewal. The questions worth asking before you sign:
What is the hardware actually worth, priced separately, and what happens to it if you change platforms?
Does the display hardware appear anywhere in your capital plan, or only in your operating expense?
Over a 5 to 7 year replacement cycle, what is the total, including the subscription and the processing spread?
Who is responsible when an outdoor display fails in February, and what is the response time commitment?
There is a middle path that gets overlooked in this conversation. The platform and the display hardware do not have to come from the same company. The Howard Company is Square's chosen drive-thru partner
and it builds systems that integrate with the major restaurant platforms rather than requiring one, which is why its digital display line is built to be technology agnostic. Buying hardware you own outright and software you can change is a different risk profile than buying both from one vendor on one contract.
Takeaway: free hardware is a financing structure, not a price. Get it priced both ways, and ask what the hardware is worth on its own, before you compare.
The Howard Company builds a fleet number from the bottom up, with this formula:
Annual fleet signage spend =
(new locations × sign package per location)
+ (total locations × annual refresh rate × refresh project cost)
+ (displays due for replacement this year × installed cost per display)
+ (total displays × annual software, support and warranty per display)
+ content labor
+ integration, setup and revalidation
+ civil, electrical and connectivity
+ break-fix reserve
Each term maps to one of the nine categories below. Two of them have published benchmarks behind them. The rest you have to supply, and the honest version of this exercise is knowing which is which.
New location packages. Exterior, interior, drive-thru, wayfinding. Confirm whether the pylon or monument sign is inside your number, since permitting and structural work often sit in a different budget.
Remodel and refresh. Driven by your franchisor's image mandate if you have one. Get the cycle length in writing, since it is the difference between an annual line and a once a decade event.
Digital hardware replacement. Displays have a service life. Budget replacement on a cycle rather than treating each failure as an exception.
Software and support, recurring. Content management subscriptions, monitoring and warranty coverage. Roughly $200 per screen per year for software, plus support and warranty. See the recurring cost section for the full table.
Content labor, outsourced or in house. Someone builds and changes the content, and it costs either an hourly rate or your own staff time. Budget it explicitly and name who owns it.
Integration. Point of sale to menu board, order confirmation, voice AI, and pricing data. Setup, any development the pairing requires, and revalidation each time the point of sale changes. See the integration section.
Civil, structural and electrical work. Footings, trenching, primary power drops, permitting and variances. Signage is a construction trade as much as it is a manufactured product. On outdoor and pylon work this can be a substantial share of the total, and it usually sits in a different budget line owned by a different person, which is exactly how it goes missing from a signage number. Get it from your general contractor and decide deliberately whether it is inside or outside the figure you are comparing to a peer's.
Connectivity. Often already inside your site network cost, but confirm rather than assume, and confirm what segment the displays sit on.
Break-fix reserve. Vehicle strikes, vandalism, power events, water intrusion, and the failures that land outside warranty. Not optional in practice, and routinely absent from plans.
The last six are where signage budgets go wrong. A display that is bought but not integrated, not connected, not maintained, not updated and not insured against the parking lot stops earning its cost in year two.
One more thing before you capitalize it
A fabricated sign structure and an outdoor digital display do not have the same useful life, and capital policies often treat them as though they do. Published service life for digital displays is commonly 3 to 5 years, and an outdoor unit works harder than an indoor one. If your capitalization policy assigns a display the same schedule as the steel it is mounted on, you can end up carrying unamortized book value on a panel that has already failed, and taking the write-down in a year nobody planned for.
This is worth a specific conversation with your own finance team about the useful life your policy assigns to outdoor displays, rather than to signage as one category. We are not accountants and this is not tax advice, but it is the kind of mismatch that surfaces in an audit rather than in a budget meeting.
The Howard Company manufactures signage, menu boards, and digital display systems for multi-location restaurant, convenience store, grocery, bank, and dispensary brands. If you want help pressure testing a 2027 signage number for your own concept, talk to a specialist.