Build vs. buy: an honest 2026 framework for hospitality data

A regional VP called us last quarter with a budget, a deck, and a problem. Forty minutes in, we told her she did not need us. She needed a packaged revenue management system, a subscription BI tool, and a half-day of training. We sent her a short list and ended the call.

We are a consulting firm. We get paid to build custom systems. We also tell roughly one in three prospects, with conviction, that they should buy software instead of hiring us. Build vs. buy is not ideology. It is fit. Most operators should buy. Some operators should build with a partner. The trick is knowing which one you are before you sign a six-figure SOW or, worse, before you sign a three-year SaaS contract that cannot hold your portfolio.

When SaaS is the right answer

If you operate one to five properties, one brand, and your F&B is in line with standard hotel category buckets, packaged software will outperform anything we could build for you. Subscription BI tools have a decade of operator feedback baked into their default reports. Off-the-shelf forecasting platforms have ingested more booking curves than any custom model you could afford to train. Packaged revenue management systems give your revenue manager a workflow she can use on day one without an analyst sitting next to her. The reporting layer your brand ships, whether it sits on top of Opera or OnQ, will already cover most of what your GM opens on a Monday.

The signals that you should buy:

One brand. The brand’s BI portal already covers 70 percent of what leadership asks for.
Five or fewer properties. The marginal cost of a custom warehouse exceeds the marginal value at that scale.
F&B that maps cleanly to banquet, restaurant, IRD, and bar. No oddball outlets, no chef’s table tasting menus driving 12 percent of revenue.
No ownership-side reporting beyond what your management agreement already templates.
One PMS environment, one POS, one CRS.

If that is you, the honest play is to spend your money on configuration, training, and adoption. A GM who actually opens the tool every Monday returns more than any custom dashboard we could deliver. Buy the software. Pay an implementation partner for two weeks to set it up properly. Skip us.

When a partner is the right answer

The math flips at portfolio scale and brand complexity. If you operate ten or more properties, run multiple brands, and the people who fund your business want reporting that does not exist in any brand portal, packaged software stops being enough. It does not get worse. It just hits a ceiling.

The signals that you should build with a partner:

Ten or more properties across two or more brands. Brand portals stop reconciling. Your CFO is doing the reconciliation in Excel.
Owner reporting that goes beyond brand templates. An asset manager who wants gross operating profit by segment by month by property, with comp set context, will not get that from any single vendor.
F&B that does not fit standard segment buckets. Catering revenue out of Tripleseat split by social vs. corporate vs. SMERF, banquet menu engineering against Toast pmix, beverage program contribution to TRevPAR. Standard category buckets do not capture it.
Three or more PMS environments, or a mid-flight brand conversion. Your data already lives in two systems that do not talk to each other.
An enterprise data warehouse already in place but not modeled. The data is in Snowflake. The dbt project has not been written. Nobody owns the semantic layer. Looker connects to raw tables and gives wrong answers.

At this scale, the VP F&B and the portfolio CFO are not asking for prettier dashboards. They are asking for definitions of revenue that survive an audit, for owner monthlies that arrive on the third business day, and for board packs that do not require a 40-hour analyst sprint at quarter-end. They are asking why ProfitSage and the brand portal disagree on F&B revenue by 1.4 percent every month. Packaged software cannot ship that, because it cannot know your portfolio’s economics. A partner can.

The middle case: hybrid

Most real portfolios live here, not at either extreme. The right answer for a 15-property, two-brand operator with one custom F&B concept and an asset manager who wants quarterly board context is almost never pure buy or pure build. It is hybrid, and the boundary is the interesting question.

Our rule: buy the boring layers, build the parts your competitors do not have. Keep the packaged revenue management system, whether that is Duetto or IDeaS or whatever your brand mandates. Keep the off-the-shelf forecasting platform. Keep Lighthouse for comp set rate intelligence. Keep the subscription BI tool for property-level operational reporting. These are commodity layers. Switching costs are real but capped. A vendor will out-invest you on these problems every year for the next decade.

Then build the parts where commodity software cannot win. Owner reporting models that reconcile to your management agreements. F&B segment intelligence that knows the difference between a corporate buyout and a social wedding at the same average check. Custom forecasts for catering pace, which no packaged forecasting tool understands because catering does not behave like transient. A dbt project on top of your enterprise data warehouse that defines revenue once and serves it to every downstream tool. These are the layers that compound. They are also the layers where the IT director, the VP F&B, and the portfolio CFO actually disagree on definitions, which is exactly why a partner is more useful than a tool.

The hybrid case is harder to scope than either extreme. It is also where most of the real money is, both spent badly and earned well.

What we would actually do

Before any vendor call, before any consulting RFP, we tell operators to answer four questions in writing. Not in a meeting. On paper.

How many distinct PMS, POS, and CRS environments do you run today, and will that number change in the next 24 months?
Does your ownership group accept brand-template owner reporting, or do they want something different, and if different, can you describe it in one paragraph?
What percentage of your F&B revenue comes from segments or outlets that do not fit standard category buckets? Under 15 percent or over 15 percent.
If you had to name the single report that takes the most analyst time at quarter-end, what is it, and what would it cost to lose it?

Score yourself. If you answered one PMS, brand templates fine, under 15 percent, and the quarter-end report is a nice-to-have, you should buy SaaS. If you answered three or more PMS, custom owner reporting required, over 15 percent oddball F&B, and the quarter-end report is mission-critical, you should hire a partner. Anywhere in between is hybrid, and the only honest next step is a scoping conversation, not a software demo.

If the four-question answer points you to packaged software, our follow-up email will be a one-paragraph short list with no invoice attached. If it points you to a partner, we will tell you whether that partner is us, or someone else we trust more for your specific shape. Either way, write us back. We would rather lose the deal than sell you the wrong thing.

Peter Hough is co-founder of Hospitality Data Solutions. Twenty years in hotel and restaurant operations, now building the systems we wished we had.