Marketing attribution connects marketing campaigns and promotions with guest reservations and visits, which helps teams understand which marketing activities, sources, channels and campaigns are driving revenue.
This article explains more about how marketing attribution works in Distil, and how best to set up your marketing campaigns to achieve accurate reporting.
What is marketing attribution and why does it matter?
Marketing attribution is the process of identifying which marketing touchpoint influenced a guest to take a desired outcome such as making a reservation, or buying a gift voucher and assigning a monetary value to it.
By linking reservations and purchases back to a specific channel or campaign, attribution data shows where marketing investment is producing results.
Without attribution, all reservations appear equal regardless of how a guest discovered the venue. With attribution, reservations are segmented by origin — for example, whether a guest arrived via a paid search campaign, an organic social post, or a direct visit to a booking page.
Attribution data supports decisions about marketing spend, channel mix, and campaign performance.
About the Distil attribution model
Distil uses a linear attribution model, which means that every touchpoint on a user’s journey to making a reservation is given an equal share of the outcome.
For example - a journey that included an email click, a paid aid click, a visit from an organic social post and an organic search visit prior to booking would be divided in 4 equal parts and 25% of the value of the reservation would be assigned to each channel. Read more about how revenue is calculated below.
The attribution window is 90 days, which means that any activity outside this window does not count towards the value generated by marketing channels.
This differs from the last click attribution model typically seen in marketing tools, which assigns 100% of the outcome value to the single touchpoint immediately prior to booking. Distil does not use this model because it unfairly excludes the impact of marketing channels earlier in a guest’s journey that also contributed to generating the reservation.
How Distil collects marketing effectiveness data
Distil collects marketing effectiveness data through two methods:
Both methods work together to build a complete picture of guest interactions and booking attribution.
Tracking
The Distil tracking code provides a real-time stream of detailed information about guest interactions with a restaurant's own website.
The tracking code is a small piece of JavaScript placed directly in the header of a website, or delivered via a third-party tag management tool such as Google Tag Manager. Once installed, it collects data about the session and the user, which remains anonymous until the guest provides their details. When guest details become known — for example, at the point of a reservation or form completion — that session is resolved into a guest profile, linking prior anonymous interactions to a known guest.
Read more about Distil Tracking here.
UTM parameters
Marketers can add UTM parameters to any links used in digital marketing campaigns to identify which website sessions and subsequent reservations were generated by each activity, platform or creative asset.
When a guest follows a tracked link to make a reservation, the UTM values in that URL are captured by Distil tracking at the point of booking.
Standard UTM parameters are collected by Distil as follows:
| UTM parameter | Name | Used for | Examples |
| utm_source | Source | The name of the platform or publication where the marketing activity happens | meta, google, dotdigital, |
| utm_medium | Medium | The format or type of marketing activity | email, ppc, story, |
| utm_campaign | Campaign | The name of the campaign or a label to group all activity in a particular phase or offer together | summer_2026, new_menu_launch, |
| utm_content | Content | Used to define individual assets such as ad creatives, formats or content pieces | blog_article, portrait_video, story_ad |
For attribution data to be captured via UTM parameters, tracked URLs must be used consistently across marketing activity. Links without UTM parameters will not carry the same level of attribution information as those with.
Read more about how to construct UTMs and how they are used to assign marketing channel information in this article.
Marketing channels that contribute to attribution
| Channel | How It Is Identified | Examples of Signals Used |
| AI Tools | Visits coming from AI assistants or AI search tools | ChatGPT, Copilot, Perplexity |
| Affiliates | Affiliate or influencer-driven traffic | Affiliate mediums, partner campaigns, Affilinet |
| Direct* | Users visiting directly with no meaningful attribution | Direct source, website source, or no tracking information |
| Display | Display advertising or offline promotional campaigns | Display ads, billboards, flyers, print campaigns |
| Traffic coming from email marketing or CRM platforms | Email mediums, or sources like Klaviyo, Braze, Mailchimp, Iterable, DotDigital, etc. | |
| Local Search | Visits from Google Maps or local business listings | Google Maps source, business profile campaigns, local SEO mediums |
| Organic Search | Unpaid search engine traffic | Google, Bing, Yahoo, DuckDuckGo, search referrers |
| Paid Search | Paid advertising on search engines | Google Ads, Bing Ads, gclid, msclkid, CPC/PPC mediums |
| Paid Social | Paid advertising on social media platforms | Facebook Ads, Instagram Ads, TikTok Ads, Meta Ads, fbclid, ad set IDs |
| Referral | Traffic from another website linking to the site | Referral mediums or matching referrer/source values |
| SMS | Visits generated from SMS campaigns | SMS medium or SMS source |
| Social | Organic traffic from social media | Facebook, Instagram, LinkedIn, TikTok, Pinterest, etc. |
*It is important to note that a Direct touchpoint is only assigned a value in the attribution model if no other discernable touch points exist. For more information, see how Direct traffic is handled below.
Marketing channels that do not contribute to attribution
| Channel | How It Is Identified | Examples of Signals Used |
| Unassigned | Some attribution exists but no channel rule matched | Unknown source/medium combinations |
| Unknown | Final fallback category | Anything not captured by earlier rules |
| None | Completely unattributed traffic | No source, medium, campaign, referrer, or click IDs present |
If you have a significant volume of unassigned touch points, speak to your account manager and ask about getting custom channel rules assigned in your attribution model to increase the accuracy of your reporting.
How Direct traffic is handled
A priority-based logic is applied to every reservation to ensure Direct doesn't mask individual marketing activity performance.
-
Real channel takes priority
If any event for a reservation has a real channel (not Direct or None), those real channels are used. Any Direct events for that reservation are converted to None and excluded.
-
Solo untracked event → Direct
If there is only one event and it has no channel, it is converted to Direct.
-
>90 days old → None
If a channel is labelled '>90 days ago' and a newer event exists, the older event is converted to None. Only the more recent signal is used.
-
Otherwise keep as-is
If none of the above conditions apply, the original channel is kept unchanged.
Important: Events assigned a channel of 'None' are never counted towards attribution. Only events with a real, resolved channel contribute to reporting.
How is revenue calculated
Distil uses three different types of revenue in its reporting. This allows for the potential revenue from reservations that have not yet been completed to still be included in all marketing effectiveness assessments.
Revenue Realised
Actual revenue from completed reservations, sourced directly from OpenTable
Revenue Assumed
Estimated revenue for future reservations, using average spend per head for the relevant restaurant and service.
Blended Revenue
When a view spans both past and future reservations, realised and assumed revenue are combined into one figure.
Comments
0 comments
Article is closed for comments.