How to Attribute Revenue to WiFi Marketing Campaigns

Why Attribution Is Hard in Hospitality

Online shops have it easier. A customer clicks an email, lands on a product page and checks out on the same site, and every step is logged against the same person.

In a venue the conversion happens at the counter. A guest reads your email on Tuesday, comes in on Thursday and pays in cash. Your email platform may record the click. Your WiFi system records the login. Your till records a sale. Three systems, and nothing that ties them together on its own.

Guest WiFi gives you the middle link: a dated visit from a person whose email address you hold. That lets you ask a narrow question with some confidence: did the people we emailed come back, and more of them than the people we did not email?

Two limits apply to every method below. A WiFi login only counts guests who connect, so anyone who comes back and stays on mobile data is invisible, which means these numbers undercount. And the login tells you nothing about spend. Revenue figures need your till data or an average spend you trust.

In VoqadoWiFi, the guest list export (a CSV you can download any time) includes each guest's email consent, number of visits, first visit date, last visit date and the venues they visited. The methods below work from that file plus your email platform's report.

Method 1: Unique Promo Code Attribution

The most precise method. Each campaign carries its own code. Staff key it into the till when a guest shows it, and your till's discount report becomes the record of which campaign produced which sale.

Advantages: Hard evidence, with spend attached. Easy to explain to anyone.

Disadvantages: It depends on staff entering codes every time, and on busy nights they will not. Guests forget to show the email. Every redemption costs you the discount, so you are paying for the measurement out of margin. Codes get shared, which inflates the count.

Best for: A campaign where you need a firm number, such as a promotion a partner is co-funding. Do not put a code in every email; guests learn to wait for one.

Method 2: Post-Campaign Return Visit Analysis

This is the method we would use for most campaigns. You hold back part of the audience, send to the rest, then compare how many from each group came back.

Setup:

1. Pick the audience, for example your At risk guests (in the VoqadoWiFi dashboard, guests who were regulars but have not visited for 30 to 89 days).
2. Export the list just before the send and split it at random into Group A and Group B. In a spreadsheet, add a column of random numbers, sort by it, and take the top half as Group A.
3. Send the campaign to Group A only. Group B gets nothing.
4. After 14 days, export again. For each group, count the guests whose last visit date falls after the send date.
5. The difference between the two return rates is the part you can credit to the campaign.

Illustration, not a benchmark: Group A has 500 guests and 18% of them come back within 14 days, which is 90 visits. Group B has 500 guests and 11% come back, which is 55 visits. The campaign is credited with 35 extra visits. If your till says an average visit is worth $28, that is $980 of revenue linked to the campaign. That is revenue, not profit, and the 18%, 11% and $28 are assumptions for the example, not figures to expect.

Advantages: No codes, no staff effort, and it counts guests who came back without bringing the email.

Disadvantages: Group B misses the email, so do not hold back guests from time sensitive messages like an event they would want to know about. With small groups, a gap of a few visits can easily be chance. Run the same test across several sends before you trust the result. The last visit column also only shows the latest visit, so a guest who came twice counts once.

Method 3: Time-Windowed Visit Uplift

The simplest method, and fine for a monthly report as long as you call it directional.

Setup:

1. Pick a recent week with no campaign as your baseline.
2. Note the number of returning guests that week from your WiFi dashboard.
3. Send the campaign and note returning guests in the following week.
4. The rise above the baseline is your rough campaign effect.

Limitations: It does not control for weather, local events, school holidays or a new menu. Compare like with like (same weekdays, no bank holiday in either week) and expect the number to move around from month to month for reasons that have nothing to do with email.

Presenting Attribution to Sceptical Stakeholders

When a partner or landlord asks how you know the email caused the visit, give them the methods in order of strength:

1. Promo code redemptions as the floor: these sales definitely came from the campaign.
2. The holdout comparison as your main estimate, with group sizes shown.
3. A plain list of what you cannot see: guests who returned without connecting to WiFi, and spend you did not record against a visit.

If you need a single conservative number, credit only guests who clicked a link in the email (not opened: Apple Mail can load the tracking pixel without the email being opened, which Mailchimp says makes open data unreliable, see Mailchimp Apple Mail Privacy Protection FAQs), who came back within 10 days of the send, and whose last visit date in your pre-send export was more than 14 days earlier. Match the click report to your guest export by email address in a spreadsheet. This deliberately undercounts, which is what makes it easy to defend.

Your next step: pick the next campaign you were going to send anyway, export the audience before it goes out, and hold back half.