You are running CTV campaigns, geofencing competitors, placing programmatic display across premium publishers, and streaming audio ads into your local market. The media is live. Impressions are being served. But when someone asks you whether it is working, whether the advertising is actually driving revenue, you do not have a clear answer. You might have click-through rates. You might have a foot traffic report from your DSP. But can you connect a specific ad impression to a specific purchase at the register? For most dispensaries, the honest answer is no.
This is the attribution problem in cannabis advertising, and it is more severe in this industry than in any other retail category. Cannabis operators cannot install Google conversion pixels, cannot run Meta CAPI integrations, and cannot use the standard e-commerce attribution stack that every other retailer takes for granted. The result is that dispensaries are spending real money on advertising without the measurement infrastructure to know what is working, what is wasting budget, and where the next dollar should go. This guide explains how to build a real attribution system for cannabis advertising, from device graphs and foot traffic measurement to POS-level media-to-sales matching.
Why Attribution Matters More in Cannabis Than Any Other Retail Category
In standard retail advertising, attribution is important but survivable without. A DTC brand running Meta ads can see purchase conversions directly in Ads Manager. A restaurant running Google Ads can track online orders through the conversion pixel. Even without perfect attribution, these businesses have reasonable visibility into what their advertising dollars are producing because the platforms they advertise on are also the platforms that track conversions.
Cannabis breaks this model completely. The channels where dispensaries advertise, CTV, programmatic display, geofencing, streaming audio, are fundamentally disconnected from the places where conversions happen, which is the physical dispensary and the point-of-sale system. There is no pixel on a CTV ad. There is no click-to-purchase path when someone sees a geofenced display ad on their phone and then walks into your store three days later. The advertising and the conversion happen in entirely different ecosystems, and without a deliberate attribution infrastructure connecting them, you are flying blind.
This matters because cannabis advertising budgets are not small. A single-location dispensary running a proper omnichannel campaign is spending $10,000 to $20,000 per month. A multi-location operator might be spending six figures monthly across markets. At those budget levels, the difference between a 3:1 ROAS and a 7:1 ROAS is the difference between a marketing program that is marginally justifiable and one that is a clear growth engine. Attribution is what tells you which one you have, and more importantly, how to shift spend toward the higher-performing channels, audiences, and creative.
The core challenge: Cannabis advertising happens on screens (CTV, mobile, desktop). Cannabis purchases happen at a physical counter. Attribution is the bridge between those two worlds, and building that bridge requires technology that most dispensaries do not know exists.
The Attribution Problem: Why Google Analytics Is Not Enough
Most dispensaries have Google Analytics installed on their website. Some have GA4 configured with event tracking. This gives them visibility into website traffic sources, page views, and maybe some basic conversion events like menu clicks or store locator usage. But for cannabis advertising attribution, Google Analytics solves approximately 10% of the problem.
Here is why. The majority of dispensary conversions are in-store visits, not online transactions. When someone sees your CTV ad on Tuesday evening, browses your Dutchie menu on their phone Wednesday morning, and walks into your store Saturday afternoon, Google Analytics sees the website visit. It does not see the CTV impression that started the journey. It does not see the in-store visit that completed it. It attributes the website session to "direct" or maybe "organic" traffic and has no knowledge that paid media initiated the entire sequence.
The second limitation is cross-device blindness. A CTV ad is served to a household's streaming device. The website visit happens on a phone. The in-store visit happens on foot. Google Analytics treats these as three unrelated events because it has no mechanism for connecting a Roku device ID to a mobile advertising ID to a physical store visit. Without a device graph that links these identifiers, the customer journey is fragmented across devices and channels with no way to reconstruct it.
Common mistake: Dispensaries that evaluate CTV or programmatic campaign performance using only Google Analytics data will systematically undercount conversions by 60-80%. GA4 can only see web sessions. It cannot see view-through conversions from CTV, foot traffic driven by geofencing, or purchases at the register connected to ad exposure. If your only measurement tool is GA4, you will kill campaigns that are actually working.
The third problem is the absence of conversion pixels. In standard digital advertising, a Facebook pixel or Google tag fires when someone completes a purchase, allowing the ad platform to connect the impression to the conversion. Cannabis dispensaries cannot place these pixels on Google or Meta because they cannot advertise on those platforms. And while cannabis-compliant DSPs do offer their own tracking pixels for website events, the pixel only captures the online portion of a journey that overwhelmingly ends offline. You need a fundamentally different measurement architecture.
How Multi-Touch Attribution Works for Dispensaries
Multi-touch attribution (MTA) is the practice of assigning conversion credit across every advertising touchpoint that a customer encountered before converting. Instead of giving 100% of the credit to the last ad someone clicked (last-click attribution) or the first ad they saw (first-touch attribution), multi-touch models distribute credit across the full journey.
For a dispensary running an omnichannel campaign, a typical customer journey might look like this:
- Day 1: Customer sees a 30-second CTV ad on Hulu for your dispensary's new product line
- Day 3: Same customer, now on their phone, is served a programmatic display ad while reading a news article, based on behavioral targeting
- Day 5: Customer drives past a competing dispensary and is captured by your geofence, receiving a mobile display ad with a first-visit promotion
- Day 6: Customer visits your website on their laptop, browses the menu, checks store hours
- Day 7: Customer walks into your dispensary and makes a $85 purchase
In a last-click model, the geofenced display ad gets 100% of the credit because it was the last ad interaction before the website visit. The CTV impression that started the entire journey gets zero credit. In a last-click world, you would logically shift all budget to geofencing and cut CTV, which would collapse the top of your funnel and eventually degrade geofencing performance too because fewer people would recognize your brand when they see the conquest ad.
Multi-touch attribution solves this by recognizing that each touchpoint played a role. Common MTA models include linear attribution (equal credit to all touchpoints), time-decay (more credit to touchpoints closer to conversion), and position-based (40% to first touch, 40% to last touch, 20% distributed across middle touchpoints). The right model depends on your campaign structure, but any multi-touch approach is dramatically better than last-click for cannabis because the customer journey spans multiple devices, multiple days, and multiple channels before ending in a physical store.
Why this matters for budget allocation: Multi-touch attribution typically reveals that CTV drives 25-35% of attributed conversions when measured properly, even though CTV generates zero clicks. Dispensaries using last-click attribution see CTV as a zero-conversion channel and cut it. Dispensaries using multi-touch attribution see CTV as their primary awareness driver and scale it. Same data, radically different conclusions.
The Device Graph: Connecting Screens to Store Visits
The device graph is the foundational technology that makes cannabis advertising attribution possible. A device graph is a database that links multiple device identifiers, such as a Roku device ID, an iPhone advertising ID, a desktop browser cookie, and a tablet identifier, to a single individual or household. Modern device graphs used in programmatic advertising contain over 700 million device profiles across the United States, covering the vast majority of connected households.
Device graphs are built through two methods: deterministic matching and probabilistic matching. Deterministic matching connects devices that share a verified common identifier, typically a login. When someone logs into Hulu on their Roku and also logs into the Hulu app on their iPhone, those two device IDs are deterministically linked because the same authenticated account was used on both. Deterministic matches are highly accurate, typically above 95% confidence.
Probabilistic matching uses behavioral signals, IP addresses, location patterns, and device characteristics to infer that multiple devices likely belong to the same person or household. If a Roku, an iPhone, and a laptop all connect to the same home WiFi network, are active during the same hours, and share similar browsing patterns, the device graph assigns a probability that they belong to the same household. Probabilistic matching extends reach significantly beyond deterministic data but operates at lower confidence levels, typically 75-85%.
Household Matching and Why It Matters for CTV
CTV advertising is served to a household device, not an individual device. The TV in your living room does not have the same advertising ID as the phone in your pocket. Without household-level device graph matching, there is no way to connect a CTV impression to any downstream action on a personal device or to a physical store visit. This is why CTV attribution requires the device graph as a prerequisite.
The attribution chain works like this: a CTV ad is served to a Roku device ID. The device graph links that Roku ID to the household and identifies the mobile device IDs associated with household members. When one of those mobile devices later appears at the dispensary's physical location (confirmed via location data), the system connects the CTV impression to the store visit. Without the device graph bridging the Roku to the phone to the location, these are three disconnected data points with no attribution value.
Foot Traffic Attribution
Foot traffic attribution measures whether people who were served your ads actually visited your dispensary. This is the single most important attribution layer for cannabis retail because the conversion event, the store visit, happens in the physical world, not on a website. Foot traffic attribution uses location data from mobile devices to detect when an ad-exposed individual enters your store.
The data infrastructure behind foot traffic attribution is substantial. Modern location intelligence platforms maintain a database of over 145 million points of interest (POI) across the United States, each mapped with precise polygon boundaries. Your dispensary is not represented as a pin on a map but as an exact building footprint polygon. This precision matters because the difference between "someone walked past your building on the sidewalk" and "someone entered your store and stayed for seven minutes" is the difference between a false positive and a real conversion.
Dwell Time Confirmation
Not every device that appears within your store's polygon boundary is a real visit. Someone walking past on the sidewalk, a delivery driver dropping off a package next door, or a device with a noisy GPS signal can all register as present within the boundary. Dwell time confirmation solves this. The attribution system requires that a device remain within the polygon for a minimum threshold, typically five minutes, before counting the visit as verified. For dispensary retail, where the average in-store visit lasts 8-15 minutes, a five-minute dwell threshold filters out virtually all false positives while capturing legitimate visits.
Verified Visits vs. Pass-By Traffic
Quality attribution platforms distinguish between three categories of location signals:
- Verified visits: The device entered the store polygon and remained for the required dwell time. This is a confirmed store visit attributed to the advertising campaign.
- Pass-by traffic: The device appeared near the store but did not enter the polygon or did not meet the dwell threshold. This is not counted as a conversion.
- Baseline traffic: Devices that visited the store but were never served an ad. This establishes the control group for measuring incremental lift, the additional visits that advertising generated above what would have occurred without any campaign.
The metric that matters is incremental foot traffic lift: the percentage increase in store visits among the ad-exposed group compared to the unexposed control group. A well-run dispensary campaign should produce an incremental lift of 15-40% depending on market competitiveness and campaign maturity. If your foot traffic report shows attributed visits but no incremental lift analysis, you are seeing correlation, not causation.
Cannabis-specific insight: Foot traffic attribution matters more for dispensaries than for almost any other retail category because online ordering represents only 15-25% of total dispensary revenue in most markets. The other 75-85% walks through the door. If you are measuring advertising performance through website conversions alone, you are missing the vast majority of the revenue your campaigns are generating.
Media-to-Sales Attribution
Foot traffic attribution tells you that advertising drove people to your store. Media-to-sales attribution tells you what they bought when they got there. This is the highest-fidelity attribution layer available to dispensaries, and it is where the true ROAS calculation lives.
Media-to-sales attribution works by integrating your point-of-sale data with the advertising attribution system. When a device that was served an ad visits your dispensary (confirmed via foot traffic attribution), the system then matches that visit against your POS transaction records to determine whether a purchase was made, what was purchased, and the transaction value. The connection typically happens through timestamp matching: if an attributed device visited the store between 2:15 PM and 2:32 PM, the system looks for POS transactions at that location within that time window.
POS Platform Integration
The major cannabis POS and CRM platforms that support attribution data connections include:
The integration process typically involves exporting anonymized transaction data (timestamp, location, transaction total, product categories) from your POS system and matching it against the attribution platform's foot traffic data. No personally identifiable customer information needs to leave your POS. The match happens on the basis of location, time window, and device graph IDs, not on customer names or phone numbers.
Alpine IQ is particularly well-positioned for this because it operates as both a CRM and an attribution data hub, allowing dispensaries to connect ad exposure data directly to customer purchase histories. If you are running Alpine IQ as your CRM and a cannabis-compliant DSP for your media, the attribution pipeline from ad impression to purchase transaction can be built as a direct integration rather than a manual data match.
Springbig and Dutchie both provide transaction-level data exports that can be fed into attribution matching systems. The key requirement is that your POS data includes accurate timestamps and location identifiers so the attribution system can match the right transaction to the right store visit to the right ad impression.
Data privacy note: Media-to-sales attribution in cannabis must be built on anonymized, aggregated data. No individual customer's name, phone number, or purchase history should be shared with advertising platforms. The attribution match uses device graph IDs and timestamp correlation, not PII. Any attribution partner that requires you to share customer-level PII from your POS is operating outside of best practices and potentially outside of state privacy regulations.
True ROAS Calculation
With media-to-sales attribution in place, the ROAS calculation becomes straightforward and trustworthy. Total attributed revenue (sum of POS transactions matched to ad-exposed store visits) divided by total media spend equals your true return on ad spend. This is not estimated. It is not modeled. It is measured from actual purchase data connected to actual ad impressions through the device graph and foot traffic pipeline.
Dispensaries with full media-to-sales attribution typically see ROAS between 4:1 and 8:1 on blended omnichannel campaigns. Geofencing conquest campaigns frequently outperform at 6:1 to 10:1 because they target active cannabis consumers at competitor locations. CTV campaigns show lower direct ROAS (2:1 to 4:1) but drive higher new customer acquisition rates, making them essential for growth even when the per-channel ROAS looks lower than conversion-focused channels.
What Your Attribution Dashboard Should Show
An attribution dashboard that actually supports decision-making needs to show more than impressions and clicks. Here are the metrics your dashboard should surface, broken down by channel and by creative where possible:
Core Dashboard KPIs
Impressions and reach tell you how many people your campaign is touching and how often. For CTV, video completion rate (VCR) matters more than click-through rate because CTV ads are non-skippable and do not generate clicks in the traditional sense. A VCR above 95% is standard for CTV; the value is in the completed view, not the click.
Click-through rate (CTR) matters for programmatic display and audio companion banners, but should not be used to evaluate CTV or geofencing performance. A dispensary that judges CTV by CTR will always conclude it is underperforming, which is a measurement error, not a media error.
Attributed foot traffic is the primary conversion metric: how many verified store visits came from people who were exposed to your ads, broken down by channel. This tells you whether CTV, geofencing, or programmatic display is driving the most physical visits.
Cost per visit (CPV) divides your media spend by attributed foot traffic visits. A strong CPV for dispensary advertising is $8-$18 depending on market. ROAS divides attributed POS revenue by media spend. CPA measures cost per new customer acquired when matched against your CRM's first-visit data.
Breakdowns That Matter
Aggregate numbers are useful for executive reporting but insufficient for optimization. Your dashboard should allow you to break down every metric by:
- Channel: CTV vs. programmatic display vs. geofencing vs. streaming audio, so you can see which channel drives the highest ROAS and shift budget accordingly
- Creative: Which ad creative, message, or offer drives the most foot traffic and the highest average transaction value at the register
- Audience segment: Behavioral segments (cannabis enthusiasts, wellness seekers, competitor visitors) to identify which audiences convert at the highest rate
- Day of week and daypart: When your ads drive the most store visits, so you can concentrate spend during high-conversion windows
- Geography: Zip-code level performance to identify which neighborhoods are responding to your campaigns and which are not
Common Attribution Mistakes Dispensaries Make
1. Relying on Last-Click Attribution Only
This is the most damaging mistake and the most common. Last-click attribution gives 100% of conversion credit to the final touchpoint before a conversion. For cannabis, where the conversion is a store visit that follows a multi-day, multi-device journey, last-click attribution systematically undervalues awareness channels (CTV, streaming audio) and overvalues the last digital touchpoint (usually a display retargeting ad or a website visit). Dispensaries that optimize based on last-click data end up cutting the top-of-funnel media that feeds their entire pipeline.
2. Ignoring CTV View-Through Conversions
CTV is a view-through medium, not a click-through medium. Nobody clicks on their television. A CTV ad that achieves a 97% video completion rate and drives a 25% incremental foot traffic lift is performing exceptionally well, but it will show zero conversions in any click-based attribution model. Dispensaries must evaluate CTV through view-through attribution, counting conversions that happen within an attribution window (typically 14-30 days) after a completed CTV view, even if no click ever occurred.
View-through vs. click-through attribution: Click-through attribution counts conversions only when someone clicks an ad and then converts. View-through attribution counts conversions when someone sees an ad (without clicking) and later converts. For CTV and most brand-awareness media, view-through is the only valid measurement model. A 14-day view-through window is standard; 30 days is appropriate for higher-consideration purchases or new market launches.
3. Not Connecting POS Data
Foot traffic attribution without POS matching tells you that advertising drove people to the store. It does not tell you whether those people actually bought anything or what they spent. A dispensary that sees 500 attributed foot traffic visits in a month knows the campaign is generating physical traffic, but without POS data, it cannot calculate ROAS, cannot measure average transaction value of ad-exposed customers vs. non-exposed customers, and cannot determine whether the advertising is profitable. Connecting your POS is what upgrades attribution from a traffic report to a revenue report.
4. Setting Attribution Windows Too Short
Attribution windows define how long after an ad exposure a conversion can be credited to that ad. A 1-day attribution window would only count store visits that happen within 24 hours of seeing an ad. For cannabis retail, where the path from first impression to store visit often spans 5-10 days, a 1-day or even 7-day window misses a significant portion of legitimate conversions. The recommended attribution windows for cannabis advertising are 14 days for click-through and 21-30 days for view-through. CTV campaigns in particular need the longer window because the medium builds intent gradually over multiple exposures.
Window length tradeoff: Longer attribution windows capture more legitimate conversions but also increase the risk of false attribution, giving credit to ads that did not actually influence the visit. The right balance depends on your campaign frequency and market size. In a dense urban market with high ad frequency, 14-day windows may be appropriate. In a suburban market with lower frequency, 30-day windows capture the slower decision cycle without inflating attribution.
Frequently Asked Questions
What is cannabis advertising attribution?
Cannabis advertising attribution is the process of connecting ad impressions across CTV, programmatic display, geofencing, and audio channels to downstream outcomes like dispensary foot traffic and point-of-sale transactions. Because cannabis dispensaries cannot use Google or Meta tracking pixels, attribution relies on device graphs, location data from 700M+ devices, and POS integrations with platforms like Alpine IQ, Springbig, or Dutchie to measure true return on ad spend.
How do dispensaries measure foot traffic from advertising?
Dispensaries measure foot traffic attribution using location data from 145M+ points of interest. When a device that was served an ad later appears at a dispensary location with a dwell time exceeding a threshold (typically 5 minutes), that visit is counted as an attributed foot traffic conversion. The system distinguishes verified visits from pass-by traffic using polygon geofences drawn to the exact building footprint and dwell time confirmation.
What is a good ROAS for dispensary advertising?
A strong ROAS for dispensary advertising ranges from 4:1 to 8:1 when measured through media-to-sales attribution connecting ad exposure to POS transactions. CTV campaigns typically show lower direct ROAS (2:1 to 4:1) but drive higher brand lift and new customer acquisition. Geofencing conquest campaigns often deliver the highest measurable ROAS (6:1 to 10:1) because they target active cannabis consumers at competitor locations. These figures require proper multi-touch attribution, not last-click measurement.
Why is last-click attribution insufficient for cannabis advertising?
Last-click attribution fails for cannabis advertising because most dispensary conversions happen in-store, not online. A customer might see a CTV ad on Tuesday, receive a geofenced display ad on Thursday, and walk into the dispensary on Saturday without ever clicking an ad. Last-click attribution would show zero conversions for that campaign. Multi-touch attribution with foot traffic and POS matching captures the full customer journey from screen to store shelf, giving credit to every touchpoint that influenced the visit.
Attribution is not a nice-to-have reporting feature for cannabis advertising. It is the infrastructure that determines whether you can make informed decisions about where your marketing dollars go. Dispensaries that build the full attribution stack, device graph matching, foot traffic measurement with dwell confirmation, and POS-level media-to-sales connection, operate with a fundamentally different level of clarity than those relying on click reports and Google Analytics. The technology exists today. The integrations with Alpine IQ, Springbig, Dutchie, and cannabis-compliant DSPs are mature. The only barrier is knowing that this system exists and making the decision to implement it. The dispensaries that do will outspend and outperform the ones that do not, because they will know exactly what their advertising is producing and where the next dollar should go.