The View From the Trenches: My CMO Listening Tour

Over the past two weeks, I've immersed myself in tech’s marketing leadership, chatting with over 50 CMOs of top tech companies across a CMO AI Club for at-Scale CMOs, GrowthX's CMO Breakfast with Marcel Santilli, Scale VC's GTM Summit and CMO Roundtable with Maria Pergolino, the AI X Conference, Dovetail’s Insight Out Conference, and several intimate CMO AI dinners, including one with Theory VC Tomasz Tunguz.

These weren't theoretical discussions but candid conversations revealing the real-world implementation of AI in marketing today. These insights come directly from the frontlines— actual adoption stories with all their hard-earned wins and honest challenges. Whether you're a CMO benchmarking your AI strategy, a marketing leader prioritizing your next initiative, or a founder reimagining your go-to-market approach, these observations offer a glimpse into how marketing's AI-powered future is unfolding.

The eight most consistent themes I encountered: AI is transforming every aspect of marketing, with smaller companies often moving faster than enterprises. There's significant tool experimentation (and thrash), heated build vs. buy debates, a surprising return to on-premises solutions, and early signs of AI replacing (not just augmenting) marketing roles. While we're still in a Wild West phase before tool consolidation, the most successful leaders are digging deep with the smartest peers, experimenting deeply, and applying their marketing expertise to guide AI implementation rather than just chasing the latest technology. Let’s dig in:

1. AI Is Transforming Every Stage of the Marketing & GTM Funnel

The examples are staggering - deep, interesting AI use and automation are everywhere from content and SEO to translation, product marketing strategy, design, lead nurture, and targeting. It has infiltrated the entire marketing funnel to SDR processes and sales.

One VC commented that he thinks Marketing is ahead of other departments in its adoption, echoing OpenAI’s comment last fall that marketing had adopted ChatGPT faster than any other team.

AI is doing work that would have been too time-consuming before, but making the actual HUMAN work on top of that knowledge even more effective. A few cool demos I’ve seen in the last few weeks include:

• A sales leader built a “grade me” button that surfaces directly in Salesforce for sales reps. It’s underpinned by CopyAI and ChatGPT to evaluate a sales rep’s performance on a call and across their account strategy. It provides near real-time feedback on how to improve their chances of winning the deal, from how they pitched to how well they have mapped out the buying committee. This allows the insightful sales leader to codify his knowledge into 24/7 support for reps - instead of requiring several hours to listen to calls, write up feedback, and debreif at a later time. It’s massively improved conversion rates. The sales leader built it with a junior technical partner in under a week.

• A superhuman technical chatbot type of product allows everyone on the buying committee to get their questions answered at any time of day without scheduling meetings. It’s smarter than the best enterprise architect, SE, or customer support rep - and can move horizontally across all the domains. It has vastly compressed the duration of sales cycles.

• A company has automated outreach from its CEO to prospects in LinkedIn, using AI to assimilate research, compose more effective, personalized notes, and do all of it at scale, maximizing daily limits within LinkedIn. The CEO, once a bottleneck, has become a meeting-driving machine, booking 3-5 additional meetings a week through this automation.

• A deep research sales and BDR prep process does a massive assessment and summary of everything about a company, including listening to earnings announcements to summarizing a target’s market context, projects, and priorities. These are mapped into a specific pitch and script for the BDRs. (Listening to earnings calls is invaluable, but previously would be too time-consuming for quick-dial BDRs. The ability to quickly get insights from a massive scale of inputs is amazing.)

Big questions remain - how will we rewrite our websites for agents? How will we train junior employees to understand more strategic concepts? How does organizational design change as these use cases expand? CMOs are navigating choppy waters.

2. The Nimble Advantage: Smaller Companies Are Leading in AI Innovation

At-scale CMOs are experimenting and deploying significant AI projects, but I’ve been impressed with many smaller companies that are running more comprehensive ai-native tech stacks, and pushing the AI boundaries harder. Often, innovation is coming out of the Growth or Demand Generation team - the hybridized technical and business skills give them more confidence and hunger to stitch together tools and automate processes.

It's a vast generalization, but I'm seeing smaller companies innovate faster - perhaps forced to adapt faster with fewer resources, perhaps having an even greater sense of urgency from the existential threat of competitors. They also don’t have the same safety and governance requirements as larger companies. (I heard of one Fortune 500 bank took an entire 12 months to get ChatGPT Enterprise approved by IT/Procurement/Governance and hesitated to pursue additional LLMs given the difficulty.)

3. The Great Tool Shuffle: Experimentation and Rapid Evolution

Many CMOs have rolled out AI tools and have seen amazing gains in productivity and conversion rates. But there’s very little loyalty. Even as a tool sees success, new improvements or requirements are creating tool thrash.

One CMO saw huge gains on an AI chatbot for top-of-funnel prospects. But she plans to swap it out at the end of the year since it doesn’t help customers with support issues well - they have to restart their process from scratch in another tool. After getting huge marketing gains, she now wants an even more comprehensive option.

4. Build vs Buy: The New Reality of Weekend AI Solutions

The CMO in the example above is looking at new horizontal chatbots, but also considering a “build it ourselves” option.

The current AI and automation tools are allowing many industrious teams to build their own solutions - sometimes over a weekend! As AI has made building less technical, from vibe coding to business users process automation, custom processes and tools are becoming increasingly sophisticated. Big companies, too, are building a lot of complex solutions in-house, wanting private LLM instances to offer more privacy protections, and believing that their custom processes will be better met by a homegrown solution instead of a vendor’s.

It’s so fast to create a minimum-viable product that confidence is high. But will they find that the difference between “pretty good” and “exceptional” is worth the cost of an outside vendor’s tools? Will constant changes in LLMs break their home-grown processes? Will they ultimately decide to focus on their core strengths instead of tool-building? We will see!

5. The On-Prem Renaissance: Security Concerns Reign

One of the funniest “everything old is new again” trends is that many vendors are seeing momentum in their “On Prem,” or products hosted locally on a customer’s own premises instead of in the cloud. After decades of moving things from private on-prem data centers to the cloud for lower cost and greater accessibility, this is an unexpected technical reversion.

But the risks of AI being hacked, manipulated, or having data extracted mean that many companies, especially in sensitive, security conscious or highly regulated industries are very-much interested in private, “on premises” implementations — some even “Air Gapped” or literally disconnected physically from everything and only accessible directly when touched.

AI translation company Lilt is seeing massive adoption of their on-prem solution in governmental and security companies, so much so, that their on-prem solution has become a differentiator for their success.

6. The Shifting Employment Landscape: From Supporting Roles to AI Replacements

The current macroeconomic pressures from tariff threats have led many companies to tighten their belts again. Some companies have paused hiring, pulled back budgets, and are cautiously planning for a tougher year than originally expected. Many companies are starting to see purchasing slowness, which again reinforces their own cutbacks.

Thus, I’m seeing a change.

Many of the CMOs I’ve spoken with over the last six months have emphasized that AI is not yet taking jobs, but is instead limiting new hiring. Especially in the area of SDR/BDRs, where AI automation is leading the charge, many CMOs have expressed that they are not cutting SDRs, but greatly increasing their goals and quotas. “You are not going to be replaced by AI, but you could be replaced by someone who is really exceptional using AI,” one CMO had coached his team.

This week, however, I started to hear more substantial stories of job cuts for AI-driven replacements. Lower-impact employees are being cut to make room for AI spend “We swap a few SDRs for an AI process building team member and tools to support them,” said one CMO.

Two other CMOs in particular have made deep cuts in their design teams ( one cutting 80%!), transferring the work to fewer high-skilled designers using AI tools. “How is the quality?” other CMOs asked. “It’s a little early to tell,” they said.

7. Tool Consolidation Is Coming, But We're Still in the Wild West

We are in a time of massive transition, with established tools and models racing to adapt, and AI-Native tools vying to become trusted.

Models are moving from an inherently rules-based to a reasoning-based approach, where the AI decides the best next step / asset / touch. This is a massive change that will either require a re-architecting of existing tools or will foster a disruption by AI-native tools.

The landscape for marketing and GTM tools has exploded, and we haven’t yet seen consolidation. Sydney Sloan, the CMO of product-rating site G2, shared details of the massive growth at the Scale VC GTM Summit

OpenAI is in the clear LLM lead, but many people use other LLMs for specific instances where they perform better. I’m hearing many people using Copy.AI and Zapier for their automation, pitting them head-to-head on cost and effectiveness. Qualified, 1Mind, and Reggie.ai are mentioned frequently by the CMOs I know in the chatbot/superhuman categories. As OpenAI continues to expand its feature set, there’s also a question of whether it will make some current AI tools obsolete.

8. Action Plan: Pragmatic Steps for Marketing Leaders in the AI Era

Like many folks, I’ve felt fear and anxiety about AI. “It’s changing everything. How do I best adapt?”

And while it is infiltrating every aspect of business, last week reminded me that it’s all built on a basis of business, marketing, and sales insights. A brilliant AI technician or prompt engineer still needs a skilled businessperson to think through what AI really needs to do. New tools also help the less technical folks implement their insights.

YOU have to understand what your customers want, how your sales process will progress, and what context is relevant to build a great process. This includes building context for prompts, fact-checking, and continuing to optimize the flow. Everything we’ve learned about marketing, sales, business, and our industry is CRITICAL (and useful) in an AI world.

Based on everything I've observed from these 50 CMOs, here are the most practical next steps you can take:

1. Focus on high-impact, time-consuming processes first. Look for activities that require (or would benefit from) synthesizing large amounts of information or involve repetitive creative tasks. Start with a pilot project that has clear ROI metrics to build momentum.

2. Create a cross-functional AI team. Include both technical and marketing talent. AI talent is hard to hire, but many of the more technical marketers in growth, revenue operations, or data may be able to step up.

3. Prioritize flexibility in your tool choices. The landscape is changing too rapidly to bet on a single platform. Choose solutions with robust APIs and integration capabilities, and negotiate shorter contract terms with vendors.

4. Map roles against AI capabilities. Identify where upskilling is urgent and invest in training programs that develop "AI collaboration skills" such as prompt engineering and process design. Restructure metrics to reward AI-enhanced productivity.

5. Personally commit to regular learning. Block time specifically for AI learning and experimentation—treat this as non-negotiable strategy time. Integrate a learning process and AI sharing into your team rituals.

6. Stay close to experts. I've booked regular coffee dates with my cutting-edge AI friends. I've created and joined several CMO clubs. I’ve joined several Marketing AI conferences – Next one up is HSE’s AI Agentic Summit, May 6-8, where I’ll be interviewing Zapier CEO Wade Foster.

7. Watch more demos. It's hard to know where the market will consolidate. I'm sitting in on as many demos as possible, finding that AI vendors are often closer to the cutting edge than in-seat marketers - they have more context and insights about what’s moving and what’s real.

8. Just do it. Where do we find the time with everything else going on? After watching some really innovative marketers this week, I'm reminded that we might need to slow down a little to speed up. Documenting processes, getting even more detailed on prompts than you’d ever imagine, hacking something out with two folks in a week sounds hard to fit in, but the results some people are getting are staggering. The journey of a thousand miles begins with one step.

Carilu Dietrich is a former CMO, most notably the head of marketing who took Atlassian public. She currently advises CEOs and CMOs of high-growth tech companies. Carilu helps leaders operationalize the chaos of scale, see around corners, and improve marketing and company performance.