The Return of Hybrid Attribution: Ad Conversion Strategies for Unusual Business Models
Ad attribution is for everyone. Yes, even you!
If you’ve ever worked in digital marketing or even published an ad online, you know that the success of your campaigns, and likely your job, depends entirely on your ability to know how many purchases came from your ads. For those who have not worked with digital marketing, this is called “ad attribution,” and it is much older and more inherent to advertising than one might expect. Systematic, measurable attribution was already well established by the early 20th century, nearly as old as modern advertising itself.
Like all sciences, the field has evolved due to a handful of simple questions: Are my efforts in vain, and if my assumptions are indeed working, what would I expect to see that would confirm it? The coupon, the free gift, and the proto-promo-code came shortly after.
In this article, I will offer a quick examination of the problems that original attribution tactics were invented to solve, some of the original methods that were traditionally used by direct-response advertisers (as informed by the legendary Claude Hopkins), how digital attribution works and how it became the only game in town, then finally why I think good old fashioned offline attribution is making a comeback and some suggestions for how you can implement it in your business, no matter how strange or unusual your industry might be.
More specifically, my argument is this: modern digital advertising made marketers expect every business model to be tracked like an e-commerce store. That works beautifully for some businesses and very badly for others. The solution, in many cases, is not to abandon digital attribution, but to combine it with older direct-response tactics that were built to solve exactly these kinds of problems.
We begin with the burning questions.
Why believe salesmen?
I, Casey, don’t have as extensive a career in marketing as many marketers you may meet, but what I lack in quantity, I make up for in quality. Or so I told myself as I was furiously reading every book on digital marketing I could find in 2023.
My previous career usually revolved around sales and marketing departments rather than existing within them, typically in the form of serving as an SME or creative lead for promotion packages they sold. When I managed a creative team at a radio station, I would regularly get called into strategy sessions with sales teams to speculate on deliverables we could reasonably add to the station pop-up event SoWs.
I distinctly remember, even in those days, being curious about the accountability of those packages the sales team sold. How did the companies that paid for the local radio station to host a contest at their storefront know that the package they bought actually drove sales? What metrics were the sales team using to justify their services? How did the sales team themselves judge the success of these events?
My creative team and I were putting all this effort into delivering digital assets for these businesses, but even then, I intuitively knew a business can’t survive on YouTube views and Facebook likes. The sales team, I theorized, must have some way of proving the profit gained from the events and the resulting assets. How else could you know if event attendees came from the radio ads, or stopped because they saw the signs from the road? If it were left up to the client to make the ROAS calculations, they could get it wrong. If you were flying blind, how could the sales team know how to price these event packages?
The question I was hovering around was that of ad attribution, and I would only get my answers on the subject six years, three career changes, and untold numbers of books later.
Eventually, though, I did get my answers. I’m giving them to you now so you can speedrun my journey.
You can thank me later.
In the beginning was the ad
The earliest forms of advertising date back to roughly 3000 BCE, in Egypt, in the form of papyrus scrolls circulated to find runaway slaves; a nice reminder that marketing always serves the social order it exists in.
Modern advertising, as we would recognize it, began alongside the invention of the printing press in the 1400s in the form of flyers that could be handed out to passersby. It really exploded in the 17th century when English newspapers began to include advertisements along with the local gossip. It’s tempting to imagine that marketing has evolved a lot from that point with the invention of the radio, the television, and the internet. However, I think on closer examination, the medium has hardly changed at all.
In a nutshell, marketing is the art of delicately suggesting and ultimately convincing a person to redirect their time, money, or resources to something they may not have considered. Though people may be thankful for the result, they are very rarely excited about the prospect of being persuaded without their explicit consent.
That insight gives us the two greatest problems in marketing: Convincing subtly, and getting a person to tell you how, where, and why they were convinced.
Today, we will focus on the latter.
With the rise of industrial automation and Taylorism in the early 20th century, philosophies of science were liberally applied to any industry they could be applied to; marketing was no exception. In his foundational book, Scientific Advertising(1923), Claude C. Hopkins explained how he applied a rigorous “hypothesis, test, theorize, optimize, repeat” feedback loop to the, at that time, largely artistic and assumption-based field. His real contribution was not inventing the desire to measure advertising, but codifying a particularly clear and influential direct-response logic for doing so.
At the time of publishing, 1923, advertisements had just begun running on the radio. Hopkins’ book, to me, represents a master of the craft attempting to convince the starry-eyed tech-evangelists of his day not to turn their back on tried and true methods so easily. Hopkins cut his teeth in direct mail advertising. In the modern framing, whereas radio would represent a one-to-many cold approach to advertising, direct mail represents a one-to-one warm approach.
What that means is radio advertisements are broadcast indiscriminately to anyone with a radio. The ads themselves have to be tailored to an audience who did not consent to hear from this specific business and who may be so populous that ad performance can be muddied and difficult to track. Or at least that was the case at the time. More on that later.
Contrast that with direct mail advertising (yes, I’m referring to those pieces of “junk mail” you probably use as kindling or ‘puppy paper’). If the mail comes from a source the prospect gave their information to willingly, they’ve already consented to material from that source and can be considered a “warm” lead rather than a “cold” lead. Additionally, ad messaging can be tailored more personally to the recipient because more is known about them prior to them receiving the ad. This second point is what allowed Hopkins to perfect his attribution science.
Hopkins used “keyed advertising” or unique codes or response channels built into the ad. This took the form of coupons with codes unique to that particular ad creative or an offer to receive a free gift if the ad was responded to in a particular way. This advancement enabled many metrics we would recognize from the digital world today, such as cost per customer and cost per dollar of sale.
Hopkins knew that human attention is a precious commodity and that the coupon itself served as a physical reminder of the ad and offer they’d encountered. A coupon for a free sample lowers the barrier to a person responding and learning more about the product, yes, but it also has the added benefit of serving as a data point on a lead generation funnel. If various ads were served to various demographics, ad success rates could be accurately attributed to the redemption of that demographic and used to make future ads better, even if that particular demographic segment had low response rates overall.
Tactics for direct response targeting weren’t limited just to coupons. Hopkins describes running ads urging people to call a number and ask to receive their free gift, but when they call to ask for “Nick,” whereas a different ad may tell people to ask for “Gill.” This era also saw the rise of the “promo code” or the “mention this ad to receive…” tactics, which were quickly adopted by radio broadcasts.
The success of these tactics illustrates a few important insights. Firstly, it is of critical importance that you have someone manning the spreadsheets. All the attribution tracking in the world amounts to nothing if the results aren’t logged. Secondly, a huge portion of the effort put into creatives and attribution tactics needs to be directed at incentivizing the prospect to report to you that they saw a particular ad, even if they are not aware that this is what they are doing.
The 1960s saw the rise of Marketing Mix Modeling and the application of modern statistics to marketing, such as using multivariate regression on sales data to optimize individual ad metrics. Test markets could be isolated city-to-city, and metrics like spend, pricing, or even local climate could be brought into mathematical models to assess ad success.
While powerful and effective, these new attribution and testing methods still had their flaws. For example, they often treated ad impressions more uniformly than reality would justify since marketers had no way to infer different levels of intent or value across responders.
And despite all its progress, that was a burden the marketing industry had to carry.
That is, until the cookie arrived.
Modern Solutions Require Modern Problems
In the mid-1990s, the internet was opened for commercial use. As with the introduction of every mass communication technology in history, its first use cases were inevitably adult entertainment and advertising; often simultaneously.
1994 marked the occurrence of the first widely recognized web banner ad. Driven by the direct response attribution precedent, the “cookie” came shortly after and enabled the hyper-specific tracking and optimization of the online marketing environment we know today.
The browser cookie, a small bit of code stored on your computer, was originally introduced to help websites remember state and recognize returning browsers. Over time, especially through third-party implementations, cookies also became a major mechanism for cross-site tracking and targeted advertising.
This advancement is hard to overstate. It represents such a qualitative leap in attribution sophistication that it is difficult to even analogize in marketing’s history up to that point. The cookie enabled advanced metrics like behavior analysis, retargeting, and, most importantly for this article, conversion tracking 2.0.
Google was one of the first to embrace and standardize digital attribution across its advertising platforms. The attribution method used in the early days of digital is known as “Last-Click Attribution.” In this model, 100% of the credit for a sale or conversion is assigned to the very last keyword, ad, or link that the customer clicked before buying. Search-era ad platforms helped normalize this kind of attribution because it was simple, legible, and easy to automate, even if it was never a perfect reflection of how persuasion works in reality.
There was obvious room for improvement. For example, customers may hear an ad on the radio, see a billboard on the freeway, or even click on an ad on social media, do nothing, and only later that same day get onto a different device and Google the offer they saw earlier. The Last-Click Attribution model misses these conversions.
As tends to be the case in the history of marketing, more advanced statistics were applied to solve this problem.
Rule-Based attribution models were designed to distribute credit across the multiple channels that a customer interacts with on their journey to making a purchase. Last-Click attribution became one attribution strategy among many. In this new paradigm, First-Click attribution could assign 100% credit to the first channel a customer interacted with, rather than the last. Linear attribution distributes credit equally across all channels a customer interacts with. Position-based attribution assigns 40% to the first click, 40% to the last click, and 20% to everything in between. Thanks to digital event timestamps, time could also be factored into the equation by giving more credit to the channel that occurred closest to the moment of purchase.
This marks the start of the modern marketing era we find ourselves in. With the rise of multi-channel and multi-touch attribution, and the internet privacy backlash that followed, the calculations and distributions of attribution credits became so sophisticated that online advertising platforms like Meta and Google created increasingly probability-based data modeling systems that attribute ad conversion rates to likelihoods based on whatever variables are available to them at the moment of purchase.
In Google’s current ecosystem, data-driven attribution has largely displaced several older rule-based defaults.
Here lies the problem. Probabilistic models, though borne out of necessity, represent a step backwards in ad attribution from certain perspectives. The variety of advertising platforms, each with its own proprietary tracking method, has complexified the online advertising space to the point of resembling casting a huge net and sorting through the catches as opposed to the direct-response paradise that the commercialized internet was promised to be.
Digital ads also entirely ignore many of the benefits of traditional advertising.
Like any good Hegelian, I think the way we solve these problems is through understanding the failings of both online and offline attribution tactics and models, and synthesizing (Hegelians don’t yell at me for using that word) them into a new method that avoids the pitfalls of either while still allowing for their benefits.
Neither online, nor offline, but a secret, much cooler third thing
Let’s assess the issue. The main problems in digital attribution stem from the separation between a customer’s digital footprint (the websites they visit, digital ads they see, the apps they use) and their physical interactions (stores they visit, phone calls they make, billboards they see). The cutting-edge marketing tactics are the ones that find creative ways to bridge that gap, and many of these tactics come straight out of 1923.
For tracking calls made from ads, marketers use rotating, dynamic phone numbers that enable intelligent routing systems to catalogue information like the physical and digital location where the call originated, the time the call was made, or even keep a log of all the keywords that were spoken by the customer on the call. That information can be used to attribute conversions to particular ads with a high degree of certainty. Modern CRM platforms can interpret this kind of data and reliably link offline conversations to individual purchases and, thus, final revenue generated. Just like in the ads of 1923, varying phone numbers allow for the manual tracking of ad attributions independently of an online platform’s probabilistic attribution model.
You can use these online platforms to distribute your ads, but that doesn’t mean you have to take their word for how your ads are performing.
Custom URLs are also representative of the latest brand of online/offline attribution tactics. QR codes make it exceptionally easy for print, television, or direct mail campaigns to capture digital identifiers for their offline customers. By creating unique URLs and landing pages for individual ads, they gain the abundant data provided by digital ads without giving up the impact, personability, or reach of traditional advertising.
A subtle advancement over the “And tell them [host of this show] sent you” commercial spots in traditional media has been the promo code. If you pay for three different podcasters to shoutout your particular brand of antimicrobial underwear on their show, you’re pretty much stuck taking their word for how many people downloaded the show and heard the ad. That means there’s no way to audit what they charged you for those shoutouts.
With the introduction of the unique promo code, for only the price of the promotional discount (typically offset by the value of the data it provides), businesses are able to track, for themselves, which purchases come from which commercials. Due to the discount or perk associated with the promo code, the customer is highly incentivised to report to you exactly which ad read caused them to make a purchase. You don’t have to take a podcaster’s word for it and can track, to the dollar, how much money those reads made you. You can then audit for yourself whether or not their pricing is fair or profitable for you.
Though slightly dystopian on face-value, the Internet of Things, mobile GPS technology, and always-on sensors have enabled autonomous offline conversion attribution.
In the United States, privacy law is a patchwork of sector-specific federal rules, FTC enforcement, platform rules, and state privacy laws. In practice, businesses often rely on notice, consent in some contexts, contractual restrictions, and opt-out rights in others. The rule of thumb for businesses, then, is to be straightforward and, when in doubt, disclose everything that you use, where you get it, how you use it, and how a person can avoid that process if they’d like.
The recent rise in cookie acceptance banners on websites coincides with the rise in the number of data collection sources people generally interact with on a day-to-day basis.
When you are asked to sign into Facebook to get onto the wifi at the mall, when you spend a certain amount of time looking at certain items before making a purchase, when you use your loyalty card, or store-specific credit card, the value of what you offer up in data far outweighs the cost of any benefits you receive for whomever extended those benefits to you.
Even if you go into a store to browse and then ultimately make your purchase on your phone later, every available piece of your digital footprint that you left behind may become a usable signal somewhere in that broader attribution chain. Some of it may sit in first-party analytics, some of it may be passed through ad and CRM integrations, and some of it may simply improve the platform’s ability to model who converted after seeing an ad.
If you saw an ad that caused you to visit the store, you likely have cookies on your device that have tracked the platform and device you saw the ad on. If that ad has a coupon or a promo code, the model gets more accurate. Your billing information at the time of purchase can be used as data points in digital ad attribution probabilistic models. Location data from your phone might allow a store’s internal CRM to match you up with physical in-store sensor information and link the demographic persona you represent to the time you spent looking at various products in various parts of the store. If you ultimately pay using a loyalty program, the model gets even better.
I bring that up to highlight what is commonly called the “Consent Barrier.” Online data collection has morphed into a trillion-dollar industry. Modern legislature is slowly but surely catching up to the technology and putting barriers in place to protect your privacy, without completely destroying said colossal industry. Location services need to be opted into, some cookie policies need to be accepted, and the cookies themselves expire automatically, etc.
What this highlights is the need for “permission marketing.”
Just to touch on this briefly, since it isn’t the focus of the article, permission marketing, popularized by Seth Godin, stresses consent-forward approaches to advertising rather than interruption-forward approaches. Again, this is a tactic straight out of 1923. Rather than an unwanted commercial that interrupts your favorite show, a pop-up ad that interrupts your internet browsing, or a sales call that interrupts your dinner, it’s far more effective to trade mass-distribution to large cold audiences for personalized, smaller distribution to consenting audiences. The “consent” in question here is not exactly the same thing as formal legal privacy consent. It is more strategic than legal.
In this sense, consent is akin to the “consent” you give to have your likeness recorded if you’re in a public place. When you sign up for a newsletter, follow someone on social media, subscribe to a podcast, or use a store’s rewards program, that is you giving implicit consent that you like this company or this person and are willing to receive more interactions from them in the future. These acts turn you from a cold prospect to a warm lead in a strategic sense, and the tactic used to get you to buy changes right along with it.
I bring up this tangential point to highlight an important consideration in online/offline hybrid ad attribution tactics: You catch more flies with honey than vinegar.
Better attribution happens when people provide you with their information; this represents a cost to them. It costs them time, and it can cost them their sense of safety and privacy. Therefore, first and foremost in these approaches, you need to be up front about how you’re using a customer’s data and, secondly, you need to give them honest, valuable reasons for providing you with additional information and permitting you to use it.
Now, if you’ve gotten to this point and you’re saying to yourself, “sure these tactics work for Walmart and Amazon, but it would never work for my industry,” I’m going to try to convince you that online/offline attribution is not only possible, but easier to do than you’d expect, no matter how strange or unique your business model is.
The only things required to make it work are a little technical understanding and a lot of creativity.
A willingness to experiment doesn’t hurt either.
Applying CaJu Creative’s “Snowflake” Attribution Model
I don’t like the negative connotation that the term “snowflake” has developed over the last decade. Businesses, like the people they consist of, are unique, each with its own story, challenges, and characteristics, wholly unique to them.
When I say “you too can do hybrid online/offline ad attribution, no matter how unusual your business is,” that’s what I mean. There is almost always a way to do it and do it right. It just takes accounting for the unique opportunities at your disposal and building a strategy for how to properly utilize them. That’s what I call my “Snowflake Attribution Model.” It’s an axiomatic belief I hold that any business can track conversions on its ads, no matter the industry, assuming they can get sufficiently creative in their tactics.
To close out this week’s edition, I’m going to play a little game. I’m going to randomly select some industries or business models that I think would traditionally have a hard time tracking ad conversions and spitball some ways in which I think those challenges could be overcome.
To start, I’ll give a couple of real examples from my clients.
Podcast Subscriber Attribution
One of my clients runs a podcast. Podcasts, as a medium built on RSS technology, were designed to be simple and comparatively resistant to the kind of platform-native user-level attribution that marketers are used to on the web.
My client wanted to run ads to promote their podcast, but a few challenges were going to make tracking their success very difficult. First, neither the distribution nor syndication platforms they were using offered Meta or Google integration, so any attribution for show subscribers would need to be attributed externally. Additionally, podcast subscribership carries no inherent dollar-value, so even if conversions could be tracked, all impressions would be treated as equal, even if some subscribers were more valuable than others.
My proposed solution was to transition them to a link hub to serve as a singular location for all of their links, principally their “subscribe now” Spotify link, that would also serve as a middle-ground between Meta/Google and Spotify. Since the particular link hub I recommended had ad platform integrations, it could serve as the destination link for ads. The subscribe button at the top could serve as the event to be optimized for.
That solved the tracking problem.
To address the economic side of the problem, I had to qualitatively reason with this client to determine what their ultimate goal was for their advertising campaigns. Ultimately, they wanted to be picked up by a network. Working backwards from there, I determined an estimated valuation their podcast would need to be worth in order to be attractive to said network. Based on current listenership, I gauged what an ad read might be worth on their show. From there, I assigned a fraction of the per-subscriber value to the conversion event. I then created a system of weighted values for various events within the link hub.
This allowed my client to not only begin advertising their podcast more effectively, but also track their progress towards being ready for pitching themselves to the network. As you can begin to see, solving attribution problems typically means picking and choosing historical attribution methods that have previously worked in one epoch or another and reutilizing them strategically based on the problems they were invented to solve. In this particular case, we created a rough kind of probabilistic attribution model with position-based characteristics.
Contact Form, Group-Rate Quote Attributions
This was a fun one.
My client offered group-rates to their customers, but gave personalized quotes for these group-rates based on submissions through a contact form on their website. Unlike purchases from their online-store, sale values from large groups were not passed through the purchase API I had set up for them. That means I was very limited in how I could optimize ads for their group-rate offerings. An additional challenge was that this particular client was hesitant to use promo codes in their advertising.
I presented them with three possible approaches, each with its own pros and cons, and let them pick from them.
The first potential solution would be building a similar estimation method to the podcast solution mentioned above. If they gave me a few examples of the quotes they gave to inquirers, I could derive an average revenue for their group rates. If they gave me an estimated conversion rate from inquiries to confirmed reservations, I could derive an estimated value for how profitable each online form submission would be. Then, using a pixel I installed on their website, I would attribute that estimated value to an ad for every form submission that ad drove. Additionally, I could weigh the value based on the number of participants in the group reservations if they added that field to the online form. This method would be the least accurate, but represented an equally minimal amount of labor to set up.
The second solution I presented involved a promo code. If they were amenable to it, I would run group rate ads, each with its own unique promo code attached to it. All they would need to do is send me the customer and sales info for every group-reservation that mentioned a promo code. That method would let me track, to the dollar, how much value the ads were bringing in, as well as test multiple creatives at the same time. This method would be the most accurate, for a small amount of labor input on their end, but would require using a promo code.
For the final option, since I knew they didn’t like promo codes generally, I outlined a hybrid online/offline attribution method we could use. If they were willing to send me sales and customer info for every single group-reservation they received, I could package this data together to align with Meta/Google offline attribution and upload it directly to the online platforms. Using data the platforms received from the contact form submissions or “call us” button, it could apply its probabilistic model to offline purchases. If the offline customer data and time of purchase aligned with pixel information from contact form submissions, then the ad that brought the user to the contact form could be credited with that purchase value. This hybrid online/offline method represented a middle ground between precision and labor input, but allowed them to do online ad attribution while not needing to run promos.
Offline Attributions for a Bar or Restaurant
Now we move on to some hypothetical situations.
Here’s how I would do online/offline ad attribution if I were a bar or a restaurant owner.
At this point, after discussing the effectiveness of promo codes, they immediately stand out as an obvious solution. The real thorn in this approach’s side would be having to rely on the bartender to keep track of which promo codes were used on which tabs or relying on waitstaff to manually pass the values of promo-applied bills to the marketing team for manual upload.
The goal here would be to automate as many of these processes as possible. So if I were in charge, I would transition my menu to a digital one if that was okay with my clientele. At the very least, I would get an e-payment system set up for order processing and bill payment. I would make sure only to work with such an e-payment service if it offered promo code functionality.
Customer orders would be processed through a computer ticketing system rather than on paper. When it came time to pay the bill, the e-readers would display a promo code entry field. That would mean my database of customer orders would now contain a feature for the specific promo code applied, where applicable.
The next trick would then be connecting these offline transactions to specific online ads. Here also, there is a “good old-fashioned” method that would work. Instituting a customer loyalty rewards program would link individual bills to individual customer identifiers like their name, email, or phone number. In addition to the promo field, a customer would also be prompted to enter their information in exchange for future discounts on repeat purchases. If the customer consented and gave me this info, I could include that data as additional features in my dataset and upload them to Meta or Google as offline attributions to identify customers who responded to individual ads.
Since all of this information would be running through an e-payment system and stored in a database, it would be relatively simple to write a program that automatically packages and uploads any promo code applied and/or loyalty program purchases directly to the online ad platforms where they could be considered for attribution.
All these systems sound expensive to implement, until you consider the value that can come from being able to accurately target customers and optimize ad spends.
Offline Attribution for Real Estate Agents
This is an extremely challenging field to optimize ads for. The main issue stems from the fact that an average real estate agent really only makes a handful of sales a year, and prospective clients can take a very long time before they reach out, even if they engage with an ad.
The solution would need to be multi-fold. Find a reasonable way to estimate the value of conversions (with weighted impressions for more engaged leads), attribute the final purchase to the ad set that brought you that client, and reliably track ad attributions even over long time periods between impression and attribution.
Here we can think back to coupons. Remember, Claude Hopkins says they serve as a physical reminder that a person saw your ad. Plenty of real estate agents hand out swag, but the real trick here would be getting someone to contact you online or offline, self-reporting how they saw your ad.
Perhaps one could offer a gift of some kind if someone responds to a particular ad and asks for a specific person. For instance, “Call this number and ask for ‘Gill’ to receive a free appraisal on your home’s value.” Alternatively, separate Google phone numbers could be registered to various ads if your name doesn’t happen to be “Gill.”
Assuming you’re diligent about tracking which leads came from which methods, this would solve the long-time-period and final sale attribution problems. The economic one would remain. Similar to the podcast link hub I outlined earlier, the average value of a sale could be determined for this particular agent. A fraction of that value could be deduced, and then manually adjusted depending on the weight of any of the advertising channels being used.
As long as each method has a reliable identifier attached to it, one could measure the success rates of various methods and adjust weights accordingly. Put QR codes on unique website landing pages in trade magazine ads, memorable promo codes on billboards, or promote a unique event or pop-up if doing a radio interview.
That would reasonably let an agent assign attribution value estimations to conversions across many different channels. Along with unique identifiers for individual ads, this would allow the optimization of advertising channels as well as individual ad creatives.
Handyperson Offline Attribution
Let’s think about this. A handyman, henceforth referred to as “handyperson,” is likely to be discovered online through local Google searches, but their conversions would mostly happen offline, or at least outside a standard e-commerce checkout flow. For example, I’ve never prepaid for or ordered a handyperson online.
If I were a handyperson, then I would begin by taking stock of exactly what channel my leads were coming from. Website form? Phone number? Some other kind of ‘contact me’ button?
At the very least, I’d need a website. Assuming that, the site’s landing pages would serve two purposes: 1. Capture lead data and 2. Store attribution parameters.
From there, the strategy mirrors some of the others we’ve talked about: unique QR codes for print media and separate Google phone numbers or email addresses for online ads. Potentially, there could be a promo code tied to a small perk or discount tied to giving additional contact info through an online form.
If you frequent trade shows, you could have different business cards with different QR codes made to test which shows are worth going to.
If you really wanted to get fancy with it, you could use a call tracking service or inform your leads that their calls would be recorded to ensure quality. Since phone calls would likely be a handyperson’s main conversion event, call tracking could be valuable in its ability to provide source attribution, call duration, caller area codes, and, as mentioned previously, a store of valuable keywords for topics and issues that are most on the minds of your customers.
Assuming they were comfortable with manual ad attribution, this handyperson could easily take stock of the data at their disposal, upload it for manual offline attribution, and use it to optimize their ad spends just the same as an e-commerce store.
And they would be comfortable if they read this newsletter.
If I were a sculptor
But then again… no.
This section has been left intentionally blank as an exercise for the reader. If you’ve made it this far, the potential approaches should be obvious.
And if they’re not, shoot me an email and let’s talk.
Until next time, stay fresh.
- Casey