The integrator's traditional position in digital signage was structural rather than contractual. Displays came from one company, players from another, the content management system from a third, and somebody had to put them in a room and make them behave. That gap was worth a margin because it was genuinely hard to close. It is now being closed by people who are not integrators, and the question every integrator principal is asking is what remains once the gap is gone.
Who is actually squeezing the AV integrator?
Not one competitor, which is what makes it awkward to answer. In August 2026 invidis set out the pressure in an analysis of markets outside Europe and North America, and the list is longer than the usual suspects. Screen manufacturers including Samsung and TCL push turnkey packages directly to customers; so do CCTV vendors and IT hardware companies. Telecommunications operators, which already bill small businesses monthly and are practised at leasing and managed services, add signage to an existing subscription. Point-of-sale platforms absorb menu boards into the restaurant system, so the customer never buys a signage network at all. Its conclusion is the sentence to take away: the market, it argues, "is not consolidating around fewer brands, but around fewer points of accountability."
None of that is predatory. A direct bundle is a reasonable product for a customer with four screens in one building who wants one invoice and one number to call. The problem for the integrator is that this used to be the entry-level work that funded the pipeline, and it is being taken by organisations with a lower cost of sale.
Why is the box-and-install margin disappearing?
Three separate mechanisms are working on it at once, and they compound.
The first is that hardware is comparable. When a buyer can put two datasheets side by side, the only remaining variable is price, and the organisation with the largest volume wins. This is why integrators increasingly specify white-label panels: not for the panel, but to make the tender uncomparable.
The second is that the player is disappearing into the display. A screen with a competent system-on-chip and a preinstalled content management system removes a line item from the bill of materials and an afternoon from the installation. Whether that is the right engineering decision for a given estate is a separate argument, covered in our comparison of a purpose-built player against general-purpose computing. Commercially, the effect is unambiguous: one fewer box the integrator sells.
The third is that the software subscription is now bought by the end customer, often directly and often before the integrator is involved. The platform vendor gets a relationship with the customer and the renewal; the integrator gets a one-off fee for making it work. That is a worse position than it looks, because the party holding the renewal is the party the customer thinks of as the supplier.
Underneath all three sits a slower change. Demand itself is fine — AVIXA's August 2025 forecast has pro-AV end-user spending rising from about $332 billion in 2025 to $402 billion by 2030, on a growth rate revised down to 3.9% from the 5.3% forecast a year earlier. The money has not left the industry. It has moved to a different layer of it.
What do customers still pay an integrator for?
The test is what a bundle cannot ship in a box. Six things survive it, and every one of them is a capability rather than a product.
- Design that survives the second site. Screen count, zoning, redundancy, bandwidth and the failure behaviour of a multi-site estate are engineering decisions taken before anything is ordered. Our reference architecture exists because most rollouts that go wrong went wrong here.
- Integration with the building. Access control, room-booking systems, sensors, fire alarm interfaces and point-of-sale data are where signage stops being decoration. A preinstalled platform on a display does not talk to a turnstile controller; somebody has to make that connection and support it.
- Content operations. The customer who bought screens to publish daily information rarely has a person free to publish daily information. Template design, feed plumbing and automation that removes the manual step is billable work that never finishes, which is the point.
- A service level someone signs. Response time, resolution time and uptime, measured and reported. Our note on signage KPIs and SLAs covers what is realistic to commit to. A vendor selling a box does not sign one of these.
- Security responsibility. Somebody has to segment the network, manage certificates, own the patch process and answer the customer's IT department. The integrator security checklist is the shortest version of that job.
- Local presence. An engineer who can be on site tomorrow, in the right time zone, speaking the right language, with spares in the country. No remote platform replicates it, and customers undervalue it until the first outage.
Read that list as a revenue plan rather than a defence. Each item is recurring or repeatable; none of them is a spread on a purchase order.
Does building your own branded CMS actually fix it?
It is the counter-move invidis identifies — white-label displays under the integrator's own brand, combined with a proprietary or customised CMS — and it is a real strategy, described far too casually.
What it buys is genuine. A branded platform is not directly comparable in a tender, which changes the conversation from unit price to fit. It lets you move margin between hardware, software and services. It puts the renewal relationship with you instead of with a vendor the customer has never met.
What it costs is usually underestimated. A branded content management system is a product commitment, not a skin. First-line support moves to you: when a schedule does not publish at seven in the morning, the customer calls the name on the console, and that name is now yours. Feature requests arrive and have to be triaged, roadmapped and explained. Release notes have to be written by someone. And the dependence on the underlying platform does not disappear — it becomes invisible to the customer while remaining entirely real to you, which is a harder position to manage, not an easier one. The consolidation happening in the CMS layer is exactly why the question of who owns the layer underneath your brand deserves an answer before you paint it.
The version of this that works is the one where the underlying build belongs to the customer or to you rather than to a shared multi-tenant platform that can be sold. That is the argument behind 123CMS: a single-tenant, customer-branded build in front of the players, engineered by the distributor and sold by the partner, so the integrator gets the branded offer without carrying the whole product team.
Where does a distributor-through-partners model fit?
It only fits if the distributor is not also a competitor, which is a thing to check rather than assume. Media La Vista sells exclusively through its authorized partner network, not directly to end users, and has operated as SpinetiX Master Distributor for the region since 2007. The split of work is stated on the partner page in plain terms: partners deploy the hardware and implement the physical layer, while the distributor designs the architecture, enforces security and consolidates top-tier technical support.
In practice that means the end-user contract, the installation and the ongoing service agreement stay with the integrator, along with the renewal. The distributor absorbs the parts that are uneconomic for one integrator to staff alone: platform engineering, escalation, the branded-CMS build, presales design. There is one deliberate exception. Under the onboarding offer, a new partner's first three projects are delivered with full technical ownership by Media La Vista, including installation and configuration inside the UAE, so that a first deployment does not fail on unfamiliarity with the platform.
What should an integrator do in the next twelve months?
A short list, in the order it is usually easiest to execute.
- Measure the split. Work out what share of last year's gross margin came from hardware resale versus design, integration, content and support. If the first number is above half, you know your exposure.
- Price the service you already give away. Most integrators are running an unpaid support contract for their top five customers. Write it down, price it, and offer it with a reported service level.
- Audit the platforms you resell. For each one: is the customer's data exportable without the vendor's console, what does the licence say happens on a change of ownership, and how long would a forced migration take.
- Pick one integration you can do that a bundle cannot. Access control, room booking, production data, queue systems — one, done properly, becomes the reason you are in the tender.
- Argue on ten-year cost, not day-one price. Bring the total cost of ownership argument into the first meeting. A 10-year player against a device replaced three times is an argument about the customer's budget.
- Decide about branding deliberately. Either commit to a branded platform with the staffing it needs, or stay a specified partner on someone else's and compete on the six capabilities above. The failure mode is doing it halfway.
The manufacturers selling direct are not making a mistake, and neither are the customers buying that way. They found the part of the job that could be standardised. What is left is the part that could not, and it is the part that renews. Talk to Media La Vista about where that leaves a specific pipeline.