Trying to attribute conversions from agent-initiated purchases is a black hole for most marketing teams. It’s a massive blind spot in performance reporting. We see so many companies that just can’t connect the dots between a digital campaign and a sale that happens over the phone, leading to completely misallocated budgets and a fuzzy picture of their actual marketing ROI. You’re basically leaving money on the table when you can’t see what’s really working.
Key Takeaways
- You have to put a custom tracking parameter in every single purchase link your agents use. This is the only way to capture specific campaign origins and get precise attribution for those off-platform sales.
- Get your CRM data talking to your marketing analytics platform. Use webhooks or an API connection to finally match up those agent-driven sales with the customer’s first digital touchpoints.
- Carve out 15% of your marketing budget just for A/B testing different calls-to-action that push people to contact an agent, think “Speak to an Expert” buttons and different placements.
- Switch to a multi-touch attribution model (like linear or time decay) so you can give proper credit to the early digital ads and content that influence a purchase an agent closes weeks later.
- Build a closed-loop reporting system that gives you weekly ROAS numbers for campaigns driving agent-assisted sales, which lets you reallocate your budget fast instead of waiting until the end of the quarter.
Teardown: “Connect to Convert” Campaign for Expe Travel
In Q3 2025, our team ran the “Connect to Convert” campaign for Expe Travel, a premium booking service that puts together bespoke itineraries. The whole point was to drive high-value bookings, which almost always need a conversation with an agent, and then correctly trace those offline conversions back to the digital ad that started it all. This was tough because Expe’s average booking is a hefty $8,500, and customers almost never drop that kind of money without talking to a human first after doing their online research.
We had a $250,000 budget to spend over 12 weeks. Our goals were a return on ad spend (ROAS) of 3.5x, keeping the cost per lead (CPL) under $50, and hitting a 10% conversion rate from that first agent contact to a final booking. The real challenge was proving our digital spend directly led to an agent’s closed deal.
Strategy: Bridging the Digital-Offline Divide
Our strategy was all about catching a user’s interest early in their trip planning and then guiding them into an agent consultation without any friction. We knew people wouldn’t just pick up the phone and call. They needed a good reason and an obvious path. So we combined targeted digital ads with a deep CRM integration to follow the customer from their very first click all the way to the final booking confirmation.
We went after affluent people, aged 35 to 65, using detailed audience segmentation on platforms like Google Ads and Meta Business Suite. These segments were built around interests like “luxury travel,” “adventure travel,” “private tours,” and “boutique hotels.” We also built lookalike audiences from Expe Travel’s existing database of high-value clients to find more people like them.
Creative Approach: Inspiration Meets Consultation
Our creative centered on aspirational travel, showing off unique places and the kind of personal service Expe provides. The ad copy hammered on the expertise of their travel advisors and the custom nature of the trips. For visuals, we used gorgeous, high-res photos and videos of exotic locations, personalized travel moments, and professional-looking agents who seemed approachable.
On display and social, we ran short 15-30 second video clips on themes like “Himalayan Treks” or “Mediterranean Yacht Charters,” which all ended with a direct call to action: “Design Your Dream Trip. Speak to an Expe Advisor.” We also used static image carousels on Meta’s platforms to show different destinations, with each one linking back to its own landing page. For search ads, we bid on long-tail keywords like “custom safari planner” and “luxury honeymoon packages Europe,” and we used ad extensions that promoted a “Free Consultation.”
Importantly, every single ad and landing page had a trackable phone number and a “Request a Call” form in a can’t-miss spot. These forms grabbed the essential lead info we needed, like their preferred callback time and what they were interested in. That data was instantly fed into Expe’s CRM, Salesforce Sales Cloud which we had already connected to our marketing analytics platform.
Targeting: Precision for High-Value Travelers
Our targeting was extremely specific. On Google Ads, we used custom intent audiences (people actively searching for specific travel terms), in-market segments for travel services and luxury goods, and detailed demographics. We also ran remarketing campaigns that hit users who looked at specific itinerary pages but didn’t fill out a form, offering them a personalized follow-up from an agent. Over on Meta Business Suite, we used interest-based targeting, uploaded custom audiences from Expe’s own customer list, and created lookalike audiences. Geographically, we focused on high-income zip codes in big cities like New York, LA, and London.
What Worked: Smooth Data Flow and Agent Empowerment
The biggest win was getting our ad platforms, landing pages, and Expe’s CRM talking to each other smoothly. When a user filled out a “Request a Call” form or dialed one of our tracked numbers, that lead was automatically assigned to an agent. The key piece of tech here was the automatic tagging of each lead inside Salesforce with the exact campaign ID, ad group, and keyword or creative that generated it. This gave agents real context for their calls and (more importantly for us) it finally let us close the attribution loop.
Once an agent converted a lead into a booking, they were trained to log the “marketing source” right there in Salesforce. An API then pulled that data back into our analytics dashboard, letting us see exactly which digital touchpoints resulted in a booked trip. For example, if someone clicked a Google Search Ad for “custom Alaska cruise” and then called an agent three days later to book, we could tie that $12,000 booking directly back to our search ad. This was a huge change from past campaigns where agent-driven sales were just logged as “direct,” making our digital efforts look far less effective than they actually were.
Our click-through rate (CTR) across all platforms ended up at 1.8%, beating our 1.2% benchmark. We served 15.2 million impressions and got 273,600 clicks. Those clicks generated 5,472 leads in the form of agent contact requests, which put our CPL at $45.68, just under our $50 target. The agents’ conversion rate from lead to booking hit 11.5%, which was better than our 10% goal.
All told, the campaign generated 630 bookings we could directly trace back to our marketing, with an average booking value of $8,800. That’s $5,544,000 in attributed revenue. On a $250,000 budget, our final ROAS was a massive 22.17x, blowing our 3.5x target out of the water. This figure showed the incredible value of setting up proper attribution for agent-initiated sales. Without this system, a huge chunk of that revenue would have been totally invisible to the marketing team.
What Didn’t Work: Initial Call Tracking Hurdles
Our call tracking was a mess at first. The initial dynamic phone number service we used had latency issues, which meant a small but significant percentage of calls weren’t being correctly tied to their source campaign. They just showed up as “direct” calls when they were really driven by our ads. We fixed this by ripping out the old provider and switching to CallRail, which gave us much more reliable real-time tracking and had better native integrations with Google Ads and Salesforce.
Another weak spot was our first batch of display ads. We targeted generic “travel” interests, and they just didn’t perform, producing a low CTR (0.7%) and a high CPL ($75). This told us that for big-ticket purchases like these, broad awareness ads are a waste of money compared to targeting users who are already showing intent or specific luxury interests. We quickly narrowed our targeting to focus only on those more qualified segments.
Optimization Steps Taken: Iterative Refinement
Over the 12 weeks, we made several key optimizations on the fly:
- Real-time Budget Reallocation: We checked performance metrics every day and shifted the budget accordingly. Campaigns with a strong ROAS, like our Google Search ads for specific luxury destinations, got more money, while we scaled back or killed underperforming segments like the broad display targeting.
- Landing Page A/B Testing: We were constantly testing page elements like headlines, button copy, and form layouts. One test of “Request a Free Consultation” versus “Speak to an Expert Advisor” found that the second version increased form submissions by 8% on desktop. A small but easy win.
- Ad Copy Refinement: We looked at which ads were generating the highest-quality leads (the ones that actually turned into bookings) and rewrote our other ads to copy those themes. We found that ads talking about “personalized service” and “exclusive access” always beat ads that just showed pretty destination pictures.
- Agent Feedback Loop: We set up a weekly check-in with Expe’s sales agents. This was invaluable. They gave us raw feedback on lead quality and common questions they were hearing, which helped us tweak our messaging. For instance, agents told us lots of leads were asking about visas, so we added a “Visa & Entry Requirements” section to the relevant landing pages.
- Multi-Touch Attribution Modeling: While we focused on the last touch before an agent call for daily optimizations, we also used a linear attribution model in GA4 to see the whole customer journey. This showed us that organic search and blog posts often played a big part in the early discovery phase, even if a paid ad got the final click. This insight helped shape our content strategy for the next quarter.
One of the most important things we learned was how much the first digital touchpoint affected the agent’s close rate. Leads that came from super-specific, long-tail search queries had a 15% higher conversion rate with agents than leads from broader social media campaigns. This just proves how critical intent-driven marketing is for high-consideration sales.
Data Analysis: The Invisible Hand of Digital
The “Connect to Convert” campaign for Expe Travel proved, without a doubt, that our digital marketing was driving agent-initiated sales even when the deal closed offline. Being able to precisely attribute these conversions completely changed how Expe understood their marketing ROI. The final cost per conversion (a full booking) was $396.83, which is an incredibly efficient number for such a high average booking value. This campaign shows that for any business with a complex sales cycle that needs a human touch, a good attribution system is essential for making smart decisions and growing the business.
When we broke it down, we saw that 70% of the bookings we could attribute to digital marketing started with a paid search ad. Another 20% came from social media ads, and the final 10% from display and programmatic. This clear breakdown gave Expe the confidence to reallocate their future marketing budgets, pushing more of their spend into high-intent search campaigns.
It just goes back to the old saying: if you can’t measure it, you can’t manage it. For too long, companies have just accepted that there’s a blurry line between digital influence and offline sales. By putting a full tracking and attribution framework in place, we turned that blurry line into a clear, measurable path to revenue. It’s not enough just to generate leads. You have to be able to trace those leads all the way to their financial outcome, especially when agents are closing the deals.
The success here wasn’t just about the final ROAS number. It was about changing how the whole organization saw marketing’s role. Before this campaign, the sales team saw digital marketing as some separate group that just worried about website traffic. Afterward, they saw marketing as a direct partner that was feeding them qualified, context-rich leads and making their jobs easier. That shift in internal perception was invaluable.
Our final ROAS of 22.17x makes a powerful case for investing in good attribution modeling, particularly when your sales funnel involves a human. It proves you can make the “invisible hand” of digital marketing completely visible and accountable. This kind of transparency justifies the marketing spend and provides the exact insights you need to keep improving.
Accurate attribution for agent-initiated sales isn’t some technical extra. It’s a foundational part of any smart marketing strategy for a high-touch sales model. Investing in solid CRM integrations and proper tracking parameters links every dollar you spend on digital campaigns directly to revenue, turning marketing from a cost center into a transparent profit driver. For more strategies, you might want to read about media buyers’ 5 strategies for 2026 success.
What is agent-initiated purchase attribution?
This is the process of tracking and giving credit to the specific marketing touchpoints that lead a customer to buy something directly from a sales agent, instead of through a self-serve checkout online.
Why is it challenging to attribute agent-initiated purchases?
The main difficulty is that the final sale happens in a totally different system (like an agent’s sales platform or just over the phone), which breaks the connection to the original digital ad or website visit. You need proper integrations between your marketing analytics and CRM to bridge that gap.
What tools are essential for attributing agent-initiated conversions?
You absolutely need a good CRM like Salesforce, a marketing analytics platform like Google Analytics 4, and call tracking software (we used CallRail). Then you need a way to tie them all together, usually with APIs or webhooks, to create a closed-loop reporting system.
How can custom tracking parameters help with attribution?
Custom tracking parameters (like UTM codes) in your URLs tag every visitor with information about where they came from, the source, medium, campaign, and even the specific ad. This data gets passed into the CRM when they become a lead, so marketing can get credit when the sale eventually closes.
What attribution model works best for agent-initiated purchases?
A “last touch before agent contact” model is great for quick feedback on what’s driving leads right now. But a multi-touch model like linear or time decay usually gives you a more honest picture by giving some credit to all the digital ads and content that influenced the customer along the way.
