Small Cell Deployment Software: Custom Build or Buy Sitetracker

What the off-the-shelf deployment products actually do well

Buy first, and consider building only when a packaged tool has visibly run out. If your node count is measured in dozens rather than hundreds, and you work one city with one pole owner, a shared tracker plus weekly discipline beats any bespoke system, because the constraint is coordination rather than software.

Sitetracker is a serious product built for deployment programmes, and it is particularly sensible where your organisation already runs it for other work. Site and lease management, document handling, task assignment, milestone reporting and vendor visibility are all present and mature. Accruent Siterra comes from the site and lease lineage and is strong on the asset and property side, which matters if your programme sits inside a tower or real estate function. Accela sits at the other end of the same transaction and is what many municipalities run to receive and process applications, so understanding it is useful whether or not you buy anything.

Do not build a document repository, a task engine or a mapping layer. Those are commodities. Do not build a construction accounting system either. What you are considering building is a narrow model of gates and dependencies, and everything around it should be bought.

There is a genuine floor too. A neutral host provider doing a handful of venue deployments a year needs a shared drive, a project manager and a checklist. Commissioning a platform at that scale replaces a spreadsheet nobody minded with software somebody has to maintain.

One more point in favour of buying, and programmes rarely weigh it properly. A packaged vendor has already met your problem in twenty other programmes, so the implementation consultant arrives with opinions about how to sequence a market entry. A custom build gives you exactly the process you described, which is only an advantage if the process you described is a good one. If your gates are currently improvised, fix the process first and buy the tool that enforces it.

Where they stop: gate sequences that differ per jurisdiction and pole owner

Here is the workflow packaged tools model badly, and it is the whole management problem in this category. A macro site is a real estate and construction exercise. A small cell programme is hundreds of individual attachments to poles you do not own, in rights of way you do not control, each needing a separate municipal application, a separate pole owner application, a separate utility power request and a separate fibre lateral.

Configuration can express a lot. What it cannot express easily is that in one city the aesthetic review runs concurrently with engineering review and either can restart the other, that in another the shot clock only starts on a complete application and the completeness determination is discretionary, and that one pole owner will not accept a power application until make ready is invoiced while the next wants them together. That is process branching, not field configuration, and the workaround is a shadow spreadsheet.

The second thing they model badly is shared dependency. Power, fibre and permits are three separate supply chains that must converge on one node around the same date. Tools track a node own status well. They handle poorly the fact that node forty waits on a fibre segment also serving nodes forty one through forty six, so one splice crew booking moves seven on air dates. Programme managers hold that in their heads, and it is the single most valuable knowledge in the operation until the person takes leave.

Here is what that looks like in practice. Candidate selected in March. Pole attachment application filed in April, returned in May for a make ready engineering study, study completed in July, charges invoiced in August. The municipal application filed in parallel in April, held in June for an aesthetic review because the node sits in a historic overlay nobody checked, redesigned with a concealment shroud in September, approved in November. Power applied for once the design froze, service point installed in January. Fibre lateral built in December, splice not completed because the vault was flooded. Node integrated fourteen months after candidate selection. Every one of those facts was known by somebody at the time. None of them was in one place.

The arithmetic: per user licensing against an amortised build

Cost this against your own programme rather than a vendor quote, and include the coordination labour, because that is where the money actually goes.

Worked example to replace with yours. Forty internal and vendor users at $150 per user per month is $72,000 a year. Add two coordinators spending roughly half their time retyping vendor weekly reports and maintaining the shadow tracker, at a loaded $85,000 each, and you carry about $157,000 annually. Across five years, $785,000.

A first release at $110,000 with year two at 18 percent runs about $190,000 over the same five years, and the marginal cost of the next market is a template rather than a licence tier. The crossover here is process variety multiplied by volume, not volume alone: roughly 100 concurrent nodes across four or more jurisdictions, or 250 concurrent nodes in two. Below fifty nodes a year in one city, the arithmetic never gets close.

One caution on licensing. Ask what a vendor charges for external vendor users, because engineering firms and construction contractors need a view and per user pricing for occasional external updaters is what pushes programmes back to email attachments, which is where the delay and the errors come from.

What a custom build actually costs

In Digital Heroes delivery experience, a first release covering the node case with jurisdiction specific gate templates, permit and pole owner application tracking with documents, and a pipeline view broken down by blocking gate runs $75,000 to $150,000 in 12 to 16 weeks. It replaces the master spreadsheet on day one. A full platform adding utility power application tracking, shared fibre and power dependency modelling with automatic forecast propagation, vendor portals, capital and purchase order tracking per node, cycle time analytics and integration with your site database runs $180,000 to $420,000 phased over 6 to 12 months.

Migration runs 10 to 25 percent of the build. Node records, applications, documents and vendor assignments move over, but the real work is reconstructing the current gate state for every live node, because that state lives in a spreadsheet, an inbox and a coordinator memory rather than in a field. Do it market by market with the coordinator who owns each one.

Year two is 15 to 20 percent annually, covering hosting, support and outside change: a municipality adopting a new portal, a pole owner revising its make ready process, a utility changing tariff options.

What pushes you up the band specifically: the count of distinct jurisdictions and pole owners, since each is a process variant rather than a row; integration with a municipal permitting system where the municipality allows it; construction drawing and photo simulation handling; and the number of external vendors needing a view.

The four situations where building wins

Regulatory fit comes first, and here it has direct commercial value. Federal shot clock rules help you only if you can prove exactly when a complete application was submitted and document everything that followed, including completeness determinations and requests for additional information. The Federal Communications Commission 2018 small wireless facility ruling set review periods of 60 days for collocation on existing structures and 90 days for new ones, and eligible facilities requests under Section 6409(a) carry their own 60 day period. Storing applications as structured records with dated correspondence, rather than as status fields, gives you a defensible timeline.

Scale economics is second, at the node and jurisdiction thresholds above.

Third is a workflow that is your competitive advantage. For a neutral host provider, deployment visibility is the product your carrier and venue customers actually buy, and a customer facing view is exactly what packaged tools handle least well. Owning that portal is worth more than owning the internal tracker.

Fourth is integration sprawl. Count them: the deployment tracker, the site or asset database, construction accounting and purchase orders, the fibre or transport planning system, municipal portals and the utility application path. When three or more must agree before you can state an on air date, the forecast is opinion.

How to decide in a week

Take a Tuesday morning and pull your fifty oldest open nodes. For each, write one line: the gate it is blocked behind and the date it entered that gate. Give yourself three hours and use only systems, not phone calls.

Then count two things. How many nodes you could not classify at all, and how many distinct blocking gates appeared across the fifty. If more than a quarter were unclassifiable, your pipeline is not visible. If more than eight distinct gates appeared, no single linear workflow will hold your programme and configuration will keep running out.

Run the second test on dependencies. Pick one fibre segment and list every node it serves. Then ask what happens to those on air dates if the splice slips three weeks. If the answer requires a person to update each node by hand, you have found the capability worth paying for, and it is the one that makes the forecast a calculation instead of a Friday afternoon estimate.

Then buy a paid discovery phase. Digital Heroes writes a signed product requirements document before any code exists, covering the gate template model for two of your busiest markets, dependency propagation, vendor access, integrations and acceptance criteria. You own that document and can take it to any firm on your shortlist. Our work is verifiable on Clutch, Trustpilot, Fiverr Vetted Pro and a D-U-N-S registration, and you meet the named engineers before signing. We are wrong for you if you want a cheap staff augmentation contract, or if you deploy thirty nodes a year and want permission to build.

Book a 30-minute call with Digital Heroes and get a written plan and a fixed quote within 48 hours.