Marketing automation’s next phase depends on connected workflows

Marketing teams can use AI tools to draft email variations, gather account intelligence and accelerate individual tasks. But those time savings do not automatically translate into stronger business results. At a November 2026 MarTech Conference session, marketing leaders argued that automation delivers lasting value when organizations connect tools to shared workflows, clear handoffs and cross-functional adoption.

The session, titled “Marketing automation is dead. Long live marketing automation!”, focused on the distance between personal experimentation and enterprise-wide use. Frans Riemersma, CEO of MartechTribe, moderated a discussion with Eric Dates, director of growth marketing at Luxury Presence; Milton Hwang, principal at Mission MarTech LLC; Morgan Lehmann, VP of global demand generation at Iterable; and Moni Oloyede, founder of MO MarTech.

Individual efficiency is only the starting point

AI-assisted work can shorten the time needed to complete a task, but panelists distinguished that improvement from optimizing an entire marketing or revenue process. A prompt that speeds up one person’s research, for example, may have little effect on the broader pipeline if the result is not incorporated into a repeatable workflow or made useful to colleagues.

Hwang cautioned that saving 20 minutes on an ad hoc task does not, by itself, improve the end-to-end pipeline. His point was that organizations need to design how work moves between people and systems, rather than treating isolated time savings as proof of operational change. For marketing leaders, that puts workflow definition alongside tool selection: teams need to know which tasks are being automated, how outputs move to the next stage and who is responsible for acting on them.

Oloyede drew a distinction between back-office automation and autonomous customer-facing activity. Automating data hygiene or pipeline maintenance presents a more manageable risk profile, she said, than allowing a system to execute customer-facing work without adequate safeguards. Regardless of a model’s capabilities, the human operator remains accountable for the outcome.

That distinction is especially relevant to email operations, where automated actions can affect customer communications. The discussion did not prescribe a specific governance framework, but its emphasis on safeguards and human accountability points to the need to consider risk before expanding automation from internal support tasks to customer-facing execution.

Shared access and handoffs matter at scale

Hwang and Lehmann described embedded platform agents as already reducing task latency by bringing account research into automation hubs and optimizing delivery timing. Those capabilities can make useful information available sooner. Their broader value, however, depends on whether teams can access and use them as part of coordinated work.

Lehmann emphasized equitable access to tools, so teams can build together rather than operate in departmental silos. Uneven access can leave groups experimenting separately, while common platforms and workflows can support more consistent collaboration. The discussion framed adoption as an organizational issue, not simply a matter of individual proficiency with AI features.

Dates warned that uncoordinated experiments can lead companies to pay multiple times for solutions addressing the same underlying need. He advocated for centralized, collaborative environments in which work is shared instead of duplicated. The concern is not only tool proliferation: fragmented experimentation can also make it harder to establish common processes and understand how different teams’ efforts fit together.

Hwang illustrated the importance of handoffs with an automated prospect dossier. Producing the dossier has limited practical value if sales partners do not know where to find it or how to act on it. The example underscores a basic requirement for automation: an output must reach the right person in a usable form, with a clear next step. Without that connection, faster generation can simply create more unused work.

Operational oversight expands beyond software seats

Scaling AI-enabled processes also adds a financial oversight challenge. Hwang noted that operational leaders need to monitor API token usage and variable consumption costs alongside traditional software-seat expenses. As organizations embed agents in workflows, those usage-based costs become another part of managing marketing technology operations.

The panel’s central message was not that automation has disappeared, but that its role is changing. Individual AI assistance can help teams move faster; organizational value depends on deliberate workflow design, responsible human oversight, shared access and functioning handoffs. For marketing leaders, the measure of progress is therefore not just how quickly a tool completes a task, but whether the surrounding operation can put the result to work without creating disconnected processes or duplicated costs.