How to choose and integrate martech to make CRM truly work.

No episode 115 from AmplificaCast, Eric Klein talk with Grazi Sbardelotto, VP and Partner of PM Webabout what's behind a martech Well chosen and well integrated. Not the catalog view of tools, not the comparison of functionalities, but what happens in practice when a company truly decides... build an operation CRM that sustains results. And the first conclusion from this episode is that martech without strategy is cost. With strategy, it's leverage.

Grazi brings to the conversation more than 20 years of experience with CRM in large companies. She started by structuring processes at a car dealership when the subject was barely covered in universities. She went on to work for technology companies, grew PM Web's CRM BU from 20 to 250 people, and today leads projects covering loyalty, conversion optimization, automation, and AI agents. The breadth of this trajectory is what gives depth to her understanding of martech.

The martech market has changed, and few companies have realized it.

One of the most practical contributions of the episode is the overview Grazi provides of how the martech market has transformed in recent years. In the past, there was a more defined competition between Exacttarget and Responsys, which was acquired by Oracle. Each went after a different type of client. And in that interval, the market changed completely.

Indian platforms like Mogage have entered the market. Turkish platforms, like Insider, which has a strong presence in Brazil. American platforms with very specific positioning, such as Braze, which has a structure developed for companies where the application is the main channel. SAP, which offers robustness for complex operations. Zoho, for different profiles. Salesforce, which continues to dominate in large accounts. The martech landscape available today is radically different from what existed ten years ago.

For companies that need to choose, this is both an opportunity and a trap. An opportunity because there is a specific solution for almost every type of need. A trap because the proliferation of martech has created the illusion that more tools mean more results. That's not the case. Results come from how martech is integrated, configured, and operated in relation to a clear objective.

Composite martech stack: the reality of large operations

Grazi is direct in describing what she finds in larger clients:

"More and more, what we're seeing is that companies have a marketing stack that's already been developed. So, before, they contracted everything from a single company, which is easier, but sometimes one company is better for one aspect, another for another, one better for CDP, and another later for CRM activation."

This multi-vendor martech landscape is the norm, not the exception. And it creates a real operational challenge: integrating everything, ensuring data flows between tools, leveraging each layer of the stack, and guaranteeing end-to-end campaign performance. Those who fail to make this martech integration work lose coherence in communication, create gaps in the database, and compromise campaign results.

PM Web deliberately chose not to be tied to any specific platform. This gave them the freedom to serve clients with pre-defined technology stacks, often decided upon outside of Brazil, without having to question the technological choice.

“We’ve become an agnostic company. This gives our clients much more confidence that we won’t push technology on them, but rather help them get the most out of that technology.”

How to evaluate the right martech for each operation

The episode doesn't offer a formula for choosing a martech solution, but it does present clear criteria that emerge from practice. Grazi raises some implicit questions throughout the conversation that any CRM manager should answer before choosing a tool.

First: what is the main channel of communication with the customer? If it's the app, a mobile-focused martech company like Braze makes more sense. If it's email, SMS, and push notifications in more traditional operations, the options are different. Second: does the customer already have any martech installed? In companies with a history of investing in technology, there are compliance, security, and implementation processes involved. Switching platforms isn't simple. Sometimes, the way forward is to deepen the use of what already exists.

Third: Is the existing martech stack scalable? Grazi mentions that PM Web assesses whether a platform will have a larger customer base in the future, which ensures greater team depth and more knowledge sharing. A niche tool may solve a specific problem, but it requires the team to specialize in a martech that may not have longevity in the market.

AI in martech: what's already moved beyond the planning stage?

The conversation about artificial intelligence in this episode is one of the most concrete on the subject as applied to martech. Grazi doesn't talk about roadmaps or promised features. She talks about ongoing projects with measurable impact.

One example she gives is of a client who operates virtually the entire CRM workflow via prompts, without writing code. The entire martech operation is now controlled by instructions in natural language, reducing dependence on technical teams and accelerating the campaign production cycle. Another example is the use of social listening agents within WPP Open, a platform of which PM Web is a part.

"We create a social listening agent who goes to the social network, understands what the brand's end customer is saying, brings it to us, who are CRM specialists, and we build campaigns based on that."

The already measurable result is a 30 to 40% gain in agility in martech processes. The team expects to reach 60 to 70% as workflows mature and agents become more precise.

The hidden cost of martech with AI.

One of the most honest points in the episode is the warning about token costs. The conversation about martech with AI often ignores this fact. Grazi doesn't ignore it. "The client sometimes doesn't realize that, look, the more AI in their process, the more tokens they'll need. Whether it's from their own LLM or a global provider like GPT, like Claude, this will have a cost."

In an internal experiment, the team put multiple agents to work simultaneously and saw the cost skyrocket. The developer had activated several agents working together, and the token consumption far exceeded expectations. This episode changed how PM Web approaches pricing AI-based martech projects.

The man-hour or dedicated FTE model starts to lose its meaning when part of the work is done by agents. “I might have to think about different ways too. I can deliver more in less time, but with fewer people, more tokens.” This is one of the most open questions in the martech market today, and few companies are discussing it seriously.

The martech professional who will remain relevant.

When the episode turns to the topic of professional profiles, Grazi is straightforward. The martech professional who will lead projects in the coming years is the one who uses AI to expand their deliverables, solves problems with or without available tools, and knows how to collaborate with agents in the same way they collaborate with people.

What will no longer have a place is the attitude of those who wait for instruction, refuse challenges, or depend on manual processes to deliver the basics. Not because AI will replace these people, but because the pace of the martech market will no longer wait for those who don't adapt. "If we always assume 'I'm going to use AI, I'm going to respond with AI, I'm not going to think about it,' forget it. Everyone will become dumber."

This statement encapsulates what many martech teams need to hear. Automation without critical thinking produces speed without direction. And speed without direction rarely generates results.

Martech is infrastructure, not a solution.

One of the most mature points in the episode is when Grazi states that martech doesn't solve the problem of those who don't know what they want to do with their clients. The tool amplifies what already exists. If there's a strategy, martech enhances it. If there's confusion, martech scales it.

This reading is especially relevant for companies investing in martech before defining their objective. Before choosing a platform, it's necessary to know which part of the CRM operation needs the most support: CDP, activation, automation, customer service, or loyalty. Each of these layers may require different martech solutions. And trying to solve everything with a single platform, most of the time, doesn't work.

Want to understand how the right martech transforms CRM operations?

watch the episode 115 from AmplificaCast with Grazi SbardelottoJoin us, VP and Partner at PM Web, for a live conversation covering everything from the emergence of CRM in Brazil to the use of AI agents in martech operations. If you lead marketing technology projects, define a tech stack for clients, or want to understand how to integrate martech intelligently and strategically, this episode gets straight to the point.

Talk to one of our experts 👇