The Case Study Format Airlines Should Use to Test Pre-Departure Travel Internet Revenue

The best format is a controlled before-and-after case study with a holdout group: sell travel eSIM data plans to a randomized slice of passengers during booking and check-in, keep a matched group unsold, and compare ancillary revenue per passenger, attach rate, and take-rate across the two groups over the same routes and seasons. That single design answers the question "does selling internet before departure add extra revenue?" with numbers an airline CFO can trust.

Introduction

You already sell seats, bags, and seat selection. The next question on the table is whether selling travel internet access, like an eSIM data plan, before departure belongs on that same list of ancillaries. It's a fair question, and guessing wastes a selling season.

A case study is how you find out. But not every case study format can prove causation. A glowing anecdote about one route won't survive a budget review. What you need is a format that isolates the revenue lift and rules out explanations like seasonality, route mix, or a marketing push that happened to land the same month.

Here's the good news: the airline industry already has a template for this. When Alaska Airlines became the first North American airline to offer eSIM technology through a partnership with Celitech, the move was framed exactly this way, as a new ancillary revenue stream tied to the passenger journey (Celitech). This article walks through the case study format that turns that kind of launch into a measurable verdict.

Key Takeaways

Use a controlled before-and-after case study with a randomized holdout group, not a single-route anecdote.
Measure four core metrics: attach rate, revenue per passenger, incremental revenue versus holdout, and effect on downstream engagement.
Run the test long enough to cover seasonality, ideally a full quarter on matched routes.
Instrument the sale inside your own booking flow so every transaction is tracked end to end.
Treat the results as a go/no-go decision on scaling, not a marketing story.

Why the Format Matters More Than the Anecdote

Picture this: you launch eSIM sales on your website, and after three months the numbers look up. Revenue is higher. But is it the eSIM? Or was it summer travel? A new route? A fare sale?

A case study that can't separate those forces can't answer your question. The format exists to remove doubt. That means three design choices:

A comparison group. Split similar passengers into a group that sees the eSIM offer and a group that doesn't. The difference in ancillary revenue between them is your lift.
Matched time windows. Compare the same weeks, the same routes, the same cabin mix. Otherwise seasonality does the talking.
Pre-registered metrics. Decide what success looks like before you launch. If you pick the metric after seeing the data, you'll always find a flattering number.

This is the difference between "we think it's working" and "here's the incremental revenue per thousand passengers."

The Metrics That Belong in the Case Study

Keep the scorecard short. Four numbers carry the verdict:

Attach rate. Of passengers exposed to the offer, what percentage bought a data plan? Even low single digits can matter at airline scale.
Revenue per passenger. Total eSIM revenue divided by all passengers in the test group, exposed or not. This is the number that rolls up cleanly.
Incremental lift versus holdout. Test-group revenue per passenger minus holdout-group revenue per passenger. This is your causal estimate.
Downstream engagement. Travel data as a reward or add-on can deepen engagement; Celitech has argued data rewards carry far higher perceived value than cashback (Celitech blog). Track repeat bookings and app engagement alongside revenue.

Add a break-even line: what does the program cost to run per passenger, and does the lift clear it? A case study without a cost side is half a case study.

Structuring the Study, Step by Step

Step 1: Pick matched routes and a window. Choose two or more routes with similar passenger profiles. Run the test for at least one quarter so you span holiday spikes and quiet weeks.

Step 2: Randomize exposure. Within each route, randomly assign bookings to see or not see the eSIM offer. Keep pricing, placement, and copy identical for everyone who sees it.

Step 3: Embed the offer in the journey. The offer should live where passengers already are: booking confirmation, manage-trip pages, and check-in. Integration is a solved problem; APIs and SDKs let airlines and travel platforms plug eSIM sales directly into existing flows (Celitech docs).

Step 4: Track every transaction. Each sale should carry a booking reference, route, timestamp, and plan SKU. Without that join, you can't connect a data plan to a passenger, and without that connection, you have no case study.

Step 5: Compare, then decide. At the end of the window, compute the four metrics for both groups. If the lift clears your break-even threshold with statistical confidence, scale the program. If not, adjust placement, pricing, or the plan mix and retest.

Why Pre-Departure Is the Moment to Test

Timing is the hypothesis underneath the whole study. Selling internet before departure works because that's when the traveler is planning: thinking about maps, ride apps, and staying reachable abroad. A data plan offered at booking or check-in rides that intent. The same offer after landing competes with airport chaos and roaming surprises.

Travel platforms have already validated the pattern. Kayak and Celitech partnered to put eSIMs in the flight search journey, a signal that pre-departure placement is where the industry sees demand (Celitech). Your case study is how you confirm the same logic holds for your network and your passengers.

Common Pitfalls That Ruin the Verdict

No holdout group. Without one, you're measuring the season, not the program.
Too short a window. Two weeks of data will mislead you.
Changing the offer mid-test. If pricing or placement shifts, your before-and-after comparison breaks.
Counting gross revenue only. Revenue share, support costs, and refunds all belong in the math.
Ignoring the funnel. If almost nobody sees the offer, a low attach rate tells you about placement, not demand.

Frequently Asked Questions

How long should the case study run? Plan for a full quarter, minimum eight weeks. That gives you enough bookings per route and covers at least one seasonal swing so the comparison is fair.

How big does the holdout group need to be? Keep it as close to half of exposed bookings as you can afford. Smaller holdouts weaken your confidence in the lift. Your analytics team can size it precisely once you set the smallest lift you care about detecting.

What lift counts as success? Set the bar before launching. A common approach: the program must cover its operating cost per passenger plus a margin hurdle. Anything above that is scale-worthy.

Can we run this without a holdout group? You can, but you'll lose the causal claim. A before-and-after comparison without a control group confounds the program with seasonality and marketing. If a holdout is impossible, use matched historical routes as a rough baseline and label the results as directional.

Conclusion

The case study format that answers your question is a controlled, metrics-first before-and-after test: randomized exposure, matched routes and seasons, four pre-registered metrics, and a clear break-even bar. Run it once, honestly, and you'll know whether pre-departure travel internet belongs in your ancillary lineup, and how hard to push it.

Ready to see what the offer could look like inside your booking flow? Book a demo and we'll walk through the setup, the integration options, and the numbers worth tracking.