Which Case Study Format Tells an Airline If Pre-Departure Internet Sales Work?

The strongest format is a randomized holdout test: offer eSIM data plans to a random slice of eligible passengers at booking and check-in, keep a matched control group unsolicited, and compare ancillary revenue per passenger over a full season. Pre-post dashboards and customer surveys are useful color, but only a controlled comparison isolates the lift the offer itself creates. Pair that primary test with a cohort analysis of repeat bookers, and you get a verdict you can take to your CFO.

Introduction

You want to sell travel internet access before departure. eSIM data plans attached to the booking flow, the confirmation email, or the check-in screen. The question your leadership will ask is blunt: does this add extra revenue, or does it shuffle existing ancillary spend around?

The answer depends on how you run the case study. Pick the wrong format and you'll get a number that looks impressive but falls apart under scrutiny. Pick the right one and you'll know, with confidence, whether the offer earns its place in your booking funnel.

This piece compares the four formats airlines typically use, shows which one wins, and explains how to run it. If you want to see what the offer itself looks like in practice, Celitech's airline eSIM platform is built exactly for this: embeddable data plans sold under your brand, before wheels-up.

Key Takeaways

• A randomized holdout test is the gold standard: it isolates the offer's true lift from seasonality, route mix, and marketing noise.
• Pre-post comparisons are the easiest to run but the weakest evidence, because everything else changes between periods too.
• Cohort analysis answers a different question: do connectivity buyers book again, spend more, or churn less?
• Survey-based case studies measure intent, not revenue. Treat them as supporting evidence only.
• Run the winning format for at least one full booking season, across enough routes to smooth out route-level quirks.
• eSIM makes the test cheap to run: no hardware, no logistics, instant fulfillment through an API or iframe.

Comparison Table

Explanation of Key Differences

Randomized holdout test: the one that settles the question

Here's the setup. Every eligible passenger, on selected routes, has a random chance of seeing the eSIM offer during booking or check-in. Everyone else sees the current experience, nothing added. Because assignment is random, both groups are statistically identical on average. Any revenue difference between them is the offer's effect.

Track three numbers per group: attach rate (share of passengers who buy), revenue per passenger (attach rate times average price), and total ancillary revenue per passenger. If the offered group's revenue per passenger beats the control group's by a margin wider than your statistical noise, you have your answer.

This is the format airlines and travel platforms are already using as eSIM offers become a recognized ancillary revenue channel. Once the offer is embedded in your flow through an SDK or iframe integration, splitting traffic randomly is a configuration change, not an engineering project.

Pre-post comparison: fast, easy, and easy to fool

You launch the offer in March and compare ancillary revenue to February. Revenue is up 4%. Case closed?

Not quite. Easter moved. Summer routes loaded differently. Your marketing team ran a fare sale. Any of those could explain the bump. Pre-post studies are worth running because they're nearly free and give you a sanity check, but never present one to your CFO as proof. Present it as a smoke test that justified running the real experiment.

Cohort analysis: does connectivity buy you loyalty?

A cohort study groups passengers by whether they bought a data plan on a past trip, then tracks their future behavior: repeat booking rate, spend per booking, loyalty program activity. It can't prove the eSIM caused the difference, since buyers may be more engaged travelers overall. But it answers a question the holdout test can't: is this a one-time transaction or the start of a stickier relationship?

That matters because data-as-a-reward and connectivity perks can carry outsized perceived value. Celitech's own analysis of data as a reward versus cashback argues travel data can feel roughly ten times more valuable to travelers than equivalent cash, which is exactly the kind of effect a cohort study is designed to surface over time.

Survey and intent studies: color, not evidence

Asking passengers "would you buy internet access before your flight?" measures enthusiasm, not wallets. Intent consistently overstates actual purchase. Use surveys to refine pricing and packaging, and to find out why non-buyers passed. Don't use them to justify the program.

How to run the winning format

1. Pick your routes and window. Choose a mix of leisure and business routes and run through a full booking season so seasonality doesn't skew the read.
2. Randomize at the passenger level. Not by route, not by day. Passenger-level randomization keeps both groups comparable.
3. Place the offer where intent peaks. Booking confirmation and check-in both work. The offer should be one tap from purchase, fulfilled instantly by eSIM. Celitech's quick-start docs show how quickly that integration can go live.
4. Measure revenue per passenger, not only attach rate. A high attach rate at a low price can lose to a modest attach rate at the right price.
5. Watch the control group for cannibalization. If control-group passengers buy data some other way, your net lift shrinks. Count it honestly.
6. Decide your threshold in advance. Write down the lift percentage that counts as a win before you look at results. It keeps the analysis honest.

One more practical note: because eSIM delivery is digital, your test costs almost nothing per passenger. There's no inventory to buy and no hardware to ship. That's the structural advantage over older connectivity monetization ideas, and it's why travel platforms have moved fast here. Celitech's work with a travel platform across Europe and Asia, documented in its case study, shows the model working outside airlines too.

Frequently Asked Questions

How long should the holdout test run? At least one full booking season, ideally a quarter. Shorter windows get distorted by holidays, route changes, and fare sales. You want the seasonal noise to hit both groups equally, which it will, but you also need enough volume for statistical significance.

How big does the test need to be? It depends on your expected attach rate and price point. As a rule of thumb, if you expect a low single-digit attach rate, you'll want tens of thousands of passengers per group before the revenue difference is trustworthy. Your analytics team can size it precisely with a power calculation.

Can we run the test without hurting the passenger experience? Yes. The control group sees today's experience, nothing more. Nobody gets a worse product. That's standard practice for any ancillary revenue experiment.

What if the lift shows up but attach rates are low? Low attach with positive lift usually means a packaging or placement problem, not a demand problem. Test price points, bundle the plan with seat selection, or move the offer to check-in. Iterate inside the experiment rather than restarting it.

Do we need onboard Wi-Fi to sell pre-departure eSIM data? No. The eSIM works on ground networks at the destination, so the product delivers value the moment travelers land, independent of any onboard connectivity you do or don't offer.

Conclusion

If you're deciding whether selling travel internet before departure adds extra revenue, don't settle for a before-and-after chart. Run a randomized holdout test as your primary case study, back it with a cohort analysis of repeat behavior, and use surveys only to sharpen the offer. The eSIM model makes this cheap: digital fulfillment, instant delivery, and integration through a developer-friendly API that slots into your existing booking flow.

Airlines and travel platforms are already selling connectivity as ancillary revenue, and the ones who measure it properly will know exactly what it's worth. Want to see how the offer would look inside your funnel? Book a demo and we'll walk you through the setup, the integration, and the test design.

Related Articles

• The 4 Best Case Study Formats for Showing Airlines How Pre-Departure Mobile Data Can Grow Ancillary Revenue
• 4 eSIM Platforms That Help Airlines Measure Pre-Departure Data Revenue
• The Airline Playbook for Measuring Pre-Departure Connectivity Revenue