Helium SEO CTO Paul DeMott on Predictive Models and Attribution for Smarter Growth
Paul DeMott, CTO of Helium SEO, shares how his agency uses proprietary predictive technology to cut wasted effort and deliver clear, measurable SEO results. In this interview, he breaks down key shifts in link building, attribution, and adapting to AI-driven search that every ecommerce brand should consider.
In this edition of the Ecommerce Authority Playbooks series, we dive into how
Helium SEO grows, retains customers, and prepares for the future of search in 2026 and beyond.
The interview
1. What’s the quick origin story of your brand, and what makes your product or positioning genuinely different from other options in your niche?
Paul DeMott: Helium SEO started because I kept watching agencies sell SEO as a black box, where clients paid monthly retainers and got vague reports with no real way to verify the work was driving results. I had already built tools like NobleSEO.io and H1seo.io out of my own frustration with guesswork-driven link building, so launching an agency built around that same predictive, data-first approach felt like the natural next step. The idea was simple: treat marketing like an engineering problem, with inputs you can measure and outputs you can test, instead of a creative discipline you just have to trust.
What makes us different is that most agencies buy generic software and layer a service on top, while we built our own attribution and backlink prediction technology from the ground up. That means when we tell a client which links are worth pursuing or which channel is actually driving conversions, it’s backed by a model we built and can explain, not a third-party dashboard we’re interpreting for them. It’s the difference between renting insight and owning it.
2. Since launch, what have been the 1-2 real turning points for your brand-specific decisions, pivots, or experiments that noticeably changed your growth or profitability-and what did you learn from them?
Paul DeMott: The real turning point was shifting our link-building process to run through the predictive backlink model instead of the manual outreach process we started with. Before that shift, our team was spending huge amounts of time pitching sites for backlinks based on surface-level metrics like domain authority, and a large percentage of that outreach produced links that never moved a client’s rankings at all. Once we rebuilt the process around predicted ranking impact, wasted outreach dropped by more than 60 percent, and client results became consistent enough that we could actually forecast timelines instead of hedging every projection.
The bigger lesson wasn’t just about the model, it was about how much of the agency’s profitability had been quietly leaking out through low-value activity that looked productive on paper. Hours spent on outreach that felt like progress but didn’t move the needle were the biggest hidden cost in the business. That experience changed how I evaluate every process now, since I ask whether an activity is measurably tied to outcome before we scale it.
3. Which 2-3 channels drive most of your revenue right now (for example SEO, paid social, email, marketplaces, influencers), and what have you learned about making those channels work in your category?
Paul DeMott: SEO and Google Ads drive the majority of our revenue, which makes sense given the industries we serve most, home services, healthcare, and manufacturing, where people are actively searching with intent rather than browsing passively. What we’ve learned about SEO in this category is that technical and entity-based optimization outperforms generic content volume, especially for local service businesses competing in tight geographic markets where being understood correctly by search engines matters more than sheer output.
With paid search, the lesson has been about attribution more than creative or targeting. Multi-location service businesses often can’t tell which campaigns are actually driving phone calls versus form fills versus walk-ins without real cross-channel tracking, so our attribution technology has become as important to that channel’s performance as the ad copy itself. Knowing which touchpoint actually closed the sale lets us reallocate budget toward what’s working instead of what simply generated the most clicks.
4. How are you thinking about search in 2026 – Google, AI assistants like ChatGPT, and other discovery platforms? What, if anything, have you changed in your content or site to stay visible as AI search grows?
Paul DeMott: I think about search now as two parallel systems that both need to be satisfied, traditional ranking algorithms and the AI models that summarize or cite sources directly in a conversational answer. The old approach of optimizing purely for keyword relevance is losing ground to a model where entities, structured data, and clear factual authority determine whether an AI assistant pulls your content into its answer at all. That’s a fundamentally different game than ranking number one on a results page, because sometimes there’s no page being ranked, just a citation or a mention.
On our own site and for clients, we’ve leaned harder into schema markup and entity-based content structuring so that both traditional crawlers and AI models can parse exactly what a page is about and who it’s from. We’ve also started tracking brand mentions inside AI-generated answers as a visibility metric, separate from click-through data, since a customer can be influenced by an AI summary without ever generating a traditional analytics event. Ignoring that shift means measuring only half of what’s actually happening to your visibility right now.
5. What do you do to turn first‑time buyers into repeat customers and advocates? Are there specific experiences, content, or community touches that work especially well for you?
Paul DeMott: The biggest lever for us has been making sure the post-purchase experience is as data-informed as the acquisition side, since most businesses spend heavily to earn the first sale and then go quiet right when the relationship actually starts. We build attribution tracking that follows a customer past the initial conversion, which lets clients see which post-purchase touchpoints, like a follow-up email sequence or a personalized offer, actually correlate with someone coming back instead of just assuming a generic newsletter is doing the job.
One specific thing that’s worked well across service-based clients is timing follow-up content around the customer’s actual usage cycle rather than an arbitrary calendar schedule, so an HVAC client gets a maintenance reminder tied to seasonal timing instead of a random 90-day email blast. That relevance is what turns a transactional buyer into someone who trusts the brand enough to refer a friend. Advocacy tends to follow naturally once someone feels like the business actually understood their timing and their problem, rather than just closing the sale and moving on.
6. If you had to write a short playbook for an ecommerce founder one stage behind you, what would you double down on over the next 12 months – and what would you stop doing entirely?
Paul DeMott: I’d tell them to double down on owned data and attribution before spending another dollar scaling paid acquisition. Most founders at that stage are pouring budget into ads without a real system for knowing which channel or touchpoint actually drove the sale, so they end up scaling the wrong thing and can’t tell why growth stalls six months later. Building that measurement layer first feels slow, but it’s the difference between spending confidently and spending on guesswork once budgets get bigger and the mistakes get more expensive.
What I’d tell them to stop doing entirely is treating SEO and paid content as separate disciplines run by separate people with no shared data. I see founders run their SEO strategy off keyword volume while their paid team runs off conversion data, and the two never talk to each other even though they’re targeting the same customer at different points in the same journey. Once we unified that at Helium, campaigns performed better simply because decisions were being made off the same evidence instead of two competing dashboards telling two different stories.
ecommerce journey and insights with Leaders Perception’s readers.
Want to share your ecommerce playbook?
If you run an online brand and would like to be featured in a future Ecommerce Authority Playbooks interview,
you can submit your story and details here. It’s 100% free and takes just a few minutes.