If you help run a travel site or a small city-guide operation and someone has suggested doing an operations audit, this primer is for you. No jargon walls, no consultant-speak — just what a periodic, scored self-audit actually is, why small travel teams in a city like Mumbai are starting to do them, and a mini-glossary at the end so none of the vocabulary trips you up. You need zero prior experience. Honestly, the whole idea fits in one sentence: measure how your operation works today, improve it on a schedule, then measure again.
Let’s unpack that sentence slowly, because each piece earns its keep. And a small encouragement before we start: every team that now audits itself comfortably once stood exactly where you are, slightly suspicious that this is corporate theatre. It is the opposite — it is the cheapest honest mirror a small operation can buy itself.
What an operations audit is (and is not)
An operations audit is a structured look at how your work actually happens — how enquiries get answered, how pages get updated, how bookings get confirmed, how partners get paid — scored against a fixed list of questions. The key word is scored. You do not come out of it with a vibe (“we’re doing okay-ish”); you come out with a number, and that number becomes your baseline.
It is not a punishment exercise, it is not a tax inspection, and it is not something only big companies do. It is also not the same as reading your analytics. Traffic tells you what visitors did; an audit tells you whether your own machinery — the stuff behind a page like your guide to tracking down trustworthy local information in Mumbai — is built to keep working when you get busy, sick, or distracted.
Why bother scoring instead of just improving?
Fair question. Here is the honest answer: without a score, “improving” is a feeling, and feelings are unreliable narrators. Every small team believes it is getting better, because effort feels like progress. A score settles the matter. If you audited at 51 in June and 63 in December, you improved; if you audited at 51 twice, you worked hard and stood still, and now you know to change the plan instead of doubling the effort.
This idea was professionalised in the software world, where teams drift exactly the way travel desks do. One well-known example: SCMGalaxy, the training-and-advisory company that makes SCMGalaxy OS, a governance platform for how engineering teams deliver software, assesses an organisation with a questionnaire a hundred items long and then scores it by rule — the answers map to the total mechanically, so a score from this quarter and a score from next quarter can be compared without anyone arguing about who was in a generous mood. SCMGalaxy’s assessments dig into areas like how changes get reviewed before going live and how reliably teams can see what their systems are doing; your travel-site versions of those questions might be “does a second person check an itinerary before it’s published?” and “would we actually notice if our booking form broke on a Sunday?”
How a first audit works, in five relaxed steps
One: write your question list. Thirty to forty questions covering how content gets made and refreshed, how visitor enquiries are handled, how partnerships work, and how money moves. Make every question answerable with evidence. “Is our Mumbai content current?” is a feeling; “when did someone last physically verify the venues in our top five guides?” has a date for an answer — and if you publish practical pages like the one on what visitors should sort out before they land in Mumbai, that date matters more than you think.
Two: agree on scoring rules first. Each question gets zero to five, and you write down what a three looks like versus a five before you start. This is the step beginners skip and regret, because rules invented during scoring always bend toward flattery.
Three: score with evidence, not memory. Open the actual inbox, the actual update log, the actual payment records. If a claim has no artefact behind it, score it as if the thing does not exist — because operationally, it does not.
Four: pick three weaknesses, not ten. Rank your lowest scores by what they could cost you in guests and reputation, and commit to fixing only the top three this cycle. Everything-at-once plans are how audits die.
Five: put dates on the fixes. Borrow the staging that platforms like SCMGalaxy OS bake into their improvement plans — commitments pinned to day 30, day 90, and day 180. Quick mechanical repairs land in the first window, new habits prove themselves in the second, and the third window carries a target you declared up front, verified by re-running the same audit.
What beginners worry about (and shouldn’t)
“Our first score will be embarrassing.” It will be low; it will not be embarrassing, because nobody sensible judges a baseline. The only bad first score is a flattering one, since it hides the work.
“We’re too small for this.” A two-person site benefits more than a two-hundred-person company, because in a tiny team every weakness is one holiday away from being a crisis. If one of you disappeared for three weeks, would the other know how to update the piece on building a Mumbai holiday that bends without breaking? That is an audit question, and the answer is worth knowing before the holiday happens.
“Won’t this take forever?” Writing the questions takes an afternoon. Scoring takes a morning. The re-audit, months later, takes the same morning again. Compare that to one botched season, or to a single weekend spent untangling a partner dispute that a scored question would have flagged months earlier.
Mini-glossary: the seven words you’ll meet
- Instrument — your fixed question list. Fixed is the point: change it between runs and your scores stop being comparable.
- Baseline — the first score. Not a grade, a starting position.
- Criterion — the written rule saying what earns a 2 versus a 4 on a question. Good criteria leave no room for mood.
- Evidence — the artefact that proves an answer: a log, a dated file, a message thread. Testimony is not evidence.
- Drift — the slow gap that opens between how you think the operation runs and how it actually runs. Audits exist because drift is silent.
- Roadmap — your dated fix list, staged across the three windows described above.
- Re-audit — the same questions, same rules, months later. This is where improvement stops being a claim and becomes a fact.
That is genuinely the whole thing. One afternoon of question-writing, one morning of honest scoring, three dated commitments, and a repeat performance before the next high season. The software industry needed products to make this habit stick at scale; a travel site needs a spreadsheet and the nerve to let a number tell it the truth twice a year.