Key Takeaways
- Top performers are nearly three times more likely to have a documented AI roadmap (59% vs. 23%), which drives their 3x advantage in conversion and pipeline impact.
- Leading teams don’t wait for perfect measurement; they act on directional signals and extend AI into orchestration-heavy tactics like events and direct mail.
Artificial intelligence (AI) has stopped being an experiment in account-based marketing, it’s now the line that separates the teams building pipeline from the teams falling behind. Nearly everyone moves faster with AI, but speed alone doesn’t pay off.
The ForgeX 2026 State of AI in ABM Benchmark Report shows that 80% of underperformers never turn that speed into results, while top performers post more than triple the conversion and pipeline gains. That gap tells you something important about where B2B marketing sits right now: the tools are widely available, but the outcomes are anything but even.
Eric Wittlake, Lead Author of the ForgeX 2026 State of AI in ABM Benchmark Report, is a practitioner who has spent years tracking how B2B marketing teams adopt new technology. In our interview, Wittlake talked about how the report documents AI roadmap separates top performers from the pack, how top performers achieve 3x better conversion and pipeline impact and how you should decide where to chase cost savings and where to chase business outcomes, especially in high-touch, orchestration-heavy tactics like events and direct mail.
Demand Gen Report (DGR): Eric, thanks for joining us today. Nearly everyone gets faster with AI, but 80% of underperformers never convert that speed into results. What do top performers do differently to turn execution speed into pipeline?
Eric Wittlake: Thanks for having me. By far the biggest difference is that top performers have a documented AI roadmap (59% vs. 23% of others), which means they took the time to identify and develop the capabilities that actually move pipeline.
DGR: Why does a documented roadmap separate the winners from the pack so sharply?
Wittlake: Companies with a road map have multiple things in their favor that separate them from the average team:
- They have an intentional process to prioritize AI capabilities based on the impact that they can have on the business.
- There is an organizational commitment of resources.
- There is an individual responsible for developing AI functionality and driving AI adoption across the team.
There is a plan to train and enable the team on new AI functionality, and to develop functionality in a way that is accessible to the entire team, not just to the builders on the team.
DGR: Top performers report 57% improvement in conversion and pipeline impact compared to 18% for others, more than 3 times the rate. What behaviors drive that gap when both groups use the same core tools?
Wittlake: Beyond the roadmap that we already discussed, I’d highlight two additional big differences. Top performers are using AI across significantly more channels, including orchestration heavy tactics like events and direct mail. The second thing is they’re not waiting for perfect measurement. The lack of ability to measure the incremental impact of adding AI to an existing program holds some companies back and limits their investment. Top performers are more confident moving forward with anecdotal and directional feedback on impact.
How Belief in AI Fuels the Conviction Flywheel
DGR: The report calls early wins a “conviction flywheel,” with 88% of top performers confident AI will improve their programs again next year versus 51% of others. How much does belief itself shape future performance?
Wittlake: I believe the hands-on experience is a huge factors here. Once you’ve had experience seeing how you can apply AI in one place, you start to see more opportunities. Identifying a first opportunity and using AI to make an improvement, if you are not already, starts that flywheel.
As AI would say, belief is also doing real work here, but it is more at the organizational level. The belief in future results is why companies are able to invest even though the return on those investments continues to be difficult to granularly measure.
DGR: AI moves the needle on cost efficiency in 6 of 8 channels, but flips to business impact in relationship-driven ones. How should you decide where to chase savings and where to chase outcomes?
Wittlake: I would not focus on exclusively cost savings or conversion rate improvements in any channel. The underlying data in the survey shows that people achieve both in nearly every channel, we just saw that the cost efficiency impact was higher in most channels. In part, it’s just still early, and moving faster is making us more cost efficient in the early stages of broad AI adoption. I expect these numbers will change if we ask the same question a year from now.
DGR: What does it take to bring AI into high-touch, orchestration-heavy tactics?
Wittlake: It takes a lot more work to replace or augment the more complex and nuanced human-led process that is happening today. For example, when you are heavily constrained on audience size by things like event venues or the cost of direct mail, the decision about who to include or not include in a program becomes significantly more important. You have to understand how that decision is being made today, capture the data and document the process used to make that decision to create an AI workflow that can replicate portions of what your team is doing today. These high-touch activities oftentimes also include channels and tactics that typically aren’t the first ones connected to our AI stack, naturally making them later tactics to adopt for most companies.
Why Top Performers Act on AI Impact Before ROI Is Provable
DGR: Half of all respondents, including 43% of top performers, can’t directly quantify AI’s ROI. Why are top performers willing to act on imperfect signals instead of waiting for clean attribution?
Wittlake: This ties into our earlier discussion of belief. Top performers adopted early because of their belief in the impact that AI would have. We have seen something similar in other investments in the past, including early adopters of marketing automation or ABM.
Even without clean attribution, marketers show the impact of AI in other ways. For example, the measured results from activities that scaled significantly with AI can show the impact. You see this in the survey results, where respondents who can’t cleanly measure attribution to AI investments are still confident in the positive impact it is having on their programs.
DGR: Top performers and underperformers face completely different barriers: integration and expertise limits for the leaders, unclear ROI and poor data quality for the rest. What does each barrier reveal about where a team really sits on the adoption curve?
Wittlake: Although each company allows different things, I generally think about this as two different groups.
More advanced teams are often limited by the company, not their own marketing team. They are pushing the boundaries of what is allowed or possible in their company and what their builders are capable of, and they are ready to do more. Often this is a sign that marketing is one of the leaders of AI adoption across the organization as well.
Others are limited by their own internal marketing environment and aren’t bumping into the limits the broader organization puts on them. These teams are often missing the effective data or context AI needs, like clear ICPs, messaging, and value props. Early tests in these teams often didn’t produce, the missing data or context led to outputs that undermined confidence. This is the other side of the flywheel, these teams are stuck in early adoption phases without a clear path to results they have confidence in.
Why Incumbent Vendors Still Win the AI Buying Decision
DGR: AI-native tools aren’t the preferred approach in any segment, and 48% of teams run a hybrid stack instead. Why do incumbent vendors hold such a decisive advantage in the AI buying decision?
Wittlake: Two factors come up the most in this discussion. First, the incumbents are already in place, they don’t need to go through a purchase cycle or onboard and integrate a new provider. Second, many companies are hesitant to commit to a new AI vendor when the AI landscape is still changing so quickly.
DGR: Proprietary builders are the only group expecting ABM tech spending to fall, with 42% projecting decreases. When does building your own AI make sense, and when does it quietly become a liability?
Wittlake: Historically companies with big budgets and internal development resources were candidates to build their own internal technology. Those same companies are still candidates to build their own proprietary AI solutions. AI is now making it possible for some of the smallest companies to also build their own tech solutions. The companies in the middle, with a larger marketing team and need for supported, robust solutions, but without engineering and support for internal marketing tools, are the ones that are least likely to take a proprietary first approach.
DGR: The report concludes that top performers and others look nearly identical in how they build their tech stack, with the real difference living in experimentation, adoption, and planning. If the tools are the same, what’s stopping your team from closing the performance gap?
Wittlake: First, I don’t want to overread into the survey results. While their bias towards proprietary, hybrid, or AI-native is the same, that doesn’t mean they have the same tool stack. One company could be very quick to enable AI functionality from incumbent vendors while another isn’t.
That said, I believe we’re seeing the same pattern we’ve consistently seen in tech for B2B marketing. Technology is only the enabler. It’s the strategy and planning that led to the purchase decision, and the change management after it is brought in, that determines how successful companies will be using their technology stack.
