Why AI is making GTM engineers more valuable

AI has made writing sales outreach almost free. Research on AI in the workplace also shows that it can provide larger productivity gains for less experienced workers, while reducing performance when people use it for tasks outside what the technology handles well. As AI takes over more of the work involved in creating messages, the valuable skill is shifting from writing the message to building the system around it.

At the same time, email providers such as Gmail and Yahoo have introduced stricter authentication requirements and spam thresholds for bulk senders. Reaching prospects effectively now depends on more than generating good copy. It requires the right data, tools, workflows, deliverability practices, and automation. Building and managing these systems is where GTM engineers play an increasingly important role.

Key facts

In this article

• What happens to outbound when a message costs nothing
• Why the value is moving from the message to the system
• What a GTM engineer builds that AI can't
• How much of the demand is real
• What to hire for
• What does this mean for the SDR team you already have?
• Where RocketDevs fits
• Conclusion
• FAQ

What happens to outbound when writing a message costs nothing?

When AI can produce a convincing sales message in seconds, companies can send far more outbound email at very little additional cost. This increases competition for attention in already crowded inboxes. Email providers have responded by tightening their requirements and rejecting messages that do not meet their standards.

Gmail's sender guidelines set strict requirements for senders that send more than 5,000 messages per day to Gmail accounts. These senders must use SPF, DKIM, and DMARC authentication. Marketing and subscribed messages must also support one-click unsubscribing. Gmail tells senders to keep their spam rate below 0.10% and avoid reaching 0.30% or higher.

Gmail has also increased enforcement. Its sender guidelines FAQ states that, from November 2025, non-compliant traffic can face temporary or permanent rejection rather than simply being filtered into spam. Gmail also states that bulk sender status does not expire. Once a domain reaches the relevant threshold, the stricter requirements continue to apply.

AI use itself has also become widespread among the people building sales and marketing technology. JetBrains' 2026 survey of more than 15,000 professional developers found that 90% used AI coding agents at work at least weekly, while 68% used them daily. This means the teams building GTM systems are using AI extensively, as are the teams building similar systems for competitors.

High usage does not necessarily mean high confidence. Stack Overflow's 2025 developer survey found that 84% of respondents were using or planning to use AI tools. However, 46% said they actively distrusted AI's accuracy, compared with 33% who said they trusted it.

For outbound sales, the effect is straightforward. If AI can create a plausible personalised email almost instantly, the cost of sending another message becomes very small. The same is true for every other company using the technology. As more messages are sent, simply producing more copy becomes less valuable. Deliverability and the ability to reach the right prospects become more important.

Yahoo has introduced similar requirements. Its sender best practices require senders to use SPF and DKIM, publish a valid DMARC policy with at least p=none, process unsubscribes within two days, and keep spam rates below 0.3%. Yahoo's enforcement of these requirements began in February 2024.

This changes the economics of outbound. Companies can use AI to produce more sales messages than ever, but sending more email does not automatically produce more results. The systems behind that email now have to manage authentication, deliverability, unsubscribes, spam rates, targeting, and volume.

The result is a shift in where the valuable work happens. AI has made producing the message cheaper. Building the infrastructure that allows those messages to reach the right people has become more important.

Why is the value moving from the message to the system?

AI is making routine work faster and reducing the difference between average and highly skilled output. At the same time, AI can make mistakes that are difficult to spot without someone who understands the task and knows when the technology should not be trusted. Research into AI at work shows both effects.

One of the clearest studies looked at 5,172 customer-support agents who were given access to a generative AI assistant. Brynjolfsson, Li and Raymond, writing in The Quarterly Journal of Economics, found that AI assistance increased productivity by 15% on average, measured by the number of issues resolved per hour. The biggest improvements came from less experienced and lower-skilled workers. They became faster and improved the quality of their work. More experienced and higher-skilled workers saw smaller gains in speed and small declines in quality.

A separate study examined 4,867 software developers at Microsoft, Accenture, and a Fortune 100 company. The Management Science paper found that developers completed 26.08% more tasks when using AI. Less experienced developers had higher adoption rates and saw larger productivity gains.

For sales teams, this has an important implication. Writing a reasonable first-touch email used to be a skill that could distinguish stronger sales representatives from average ones. If AI can help almost anyone produce a reasonable message quickly, that difference becomes smaller. The value of the message itself therefore decreases.

The value does not disappear. It moves towards the systems that determine how, when, and where the message is used.

This is particularly important because AI does not perform equally well on every type of task. Dell'Acqua, McFowland, Mollick, Lifshitz and Kellogg tested this in a preregistered experiment involving 758 Boston Consulting Group consultants. Their findings were published in Organization Science.

Within what the researchers call the "jagged technological frontier", consultants using GPT-4 completed 12.2% more tasks. They also completed the tasks 25.1% faster on average, while producing work that researchers rated as significantly higher quality. However, on a task deliberately placed outside the AI system's effective frontier, consultants using AI were 19% less likely to reach the correct answer than consultants who did not use it.

This distinction matters for outbound sales. Generating a sales message is a task AI can handle relatively well. Other parts of the process are more complex. A system needs to determine whether an account is actually in the market, whether enrichment data is accurate, and whether a qualification model is incorrectly classifying a segment. It also needs to monitor whether email activity is approaching a provider's spam threshold.

These are system-level problems. They require someone who can connect the different tools and data sources, understand how the workflow operates, and recognise when AI is producing an unreliable result. That is where the role of the GTM engineer becomes more valuable.

Bloomberry's analysis of GTM engineering job postings also points to AI as a possible factor behind the role's growth. Henley Wing Chiu, who conducted the analysis, concluded that the rise of AI was likely a major factor. This is an interpretation rather than a direct measurement of causation, so it should be treated as a possible explanation rather than established fact.

The broader research provides a mechanism that makes this explanation plausible. As AI reduces the cost of producing routine sales content, the harder and more valuable work shifts towards designing, connecting, monitoring, and improving the systems around that content.

What does a GTM engineer actually build that AI can't?

A GTM engineer builds the system around the sales message. This system determines which accounts should be contacted, what data should be used, which sending domain should deliver the email, how much email can safely be sent, and how the results influence future outreach.

AI can write the email. It does not manage all of the infrastructure that determines whether the email should be sent in the first place. That infrastructure can be viewed as an outbound system with five stages. Each stage can fail without producing an obvious error.

Signals and data sourcing.

A GTM engineer defines the signals that indicate an account may be worth contacting. These could include a new hiring campaign, funding, a technology change, or a change in leadership. This logic may live in code or a workflow automation tool. If a data source changes its format, the workflow can break without generating an obvious error. The result may simply be that the system starts identifying the wrong accounts.

Enrichment.

Contact information often comes from several data providers. These providers can be connected through a waterfall, where one provider attempts to fill gaps left by another. Match rates can change over time, and different providers may return conflicting information. An incorrect email address can result in a hard bounce, which can contribute to problems with the sending domain.

Qualification.

AI can increasingly assess whether a company or contact fits a particular target profile. It might analyse a company's website or a person's professional profile and assign a qualification. The challenge is that AI can produce a confident answer even when the classification is wrong. Research on AI's limitations suggests that people can also be more likely to accept incorrect AI-generated answers on tasks that fall outside its effective capabilities. A GTM engineer therefore needs to monitor the output, measure errors across different segments, and adjust the rules or prompts when performance changes.

Sending infrastructure.

Email authentication and deliverability have made this a critical part of outbound. Sending domains need appropriate SPF, DKIM, and DMARC configuration. One-click unsubscribe also needs to be implemented correctly. RFC 8058 explains that email software can sometimes fetch URLs contained in email headers and accidentally trigger an unsubscribe, which is why the standard specifies how one-click unsubscribe requests should work.

The system must also process unsubscribes within Yahoo's required timeframe and continuously monitor spam rates against Gmail's thresholds. These controls need to operate at the domain level rather than being treated as a one-time setup.

Feedback.

Replies, bounces, and spam complaints provide information about how the system is performing. That information can then be used to change targeting and messaging. Without this feedback loop, the system can continue making the same mistakes while increasing the number of people affected by them.

The tools used to build these systems also show how much of GTM work now involves software. The open-source workflow automation platform n8n, which is commonly used to connect different tools and automate workflows, had 204,143 GitHub stars and 60,650 forks according to its GitHub API data. HubSpot's official Python client had 3,158,064 downloads from PyPI in the previous month.

These tools are examples of how companies are building software around their go-to-market operations. The work is not limited to using a CRM or sending emails manually. Increasingly, teams are connecting data, automation, AI, and sales systems into workflows that need to be maintained.

None of these tasks is particularly glamorous. That is exactly why they matter. AI has made producing the sales message much cheaper. The engineering work that determines whether that message reaches the right person, at the right time, through the right infrastructure, is becoming more important.

How much of the demand for GTM engineers is real?

There is evidence that demand for GTM engineering is growing, but the job title itself is newer than much of the work it describes. One analysis found that GTM engineering job postings increased by 205% year on year. The same analysis found that nine out of ten responsibilities listed in GTM engineer postings also appeared in RevOps engineer postings. This suggests that while the GTM engineer title is gaining attention, many of the underlying responsibilities are already established in revenue operations.

It is also important to understand where many of the commonly quoted figures about GTM engineering come from. Bloomberry's analysis examined GTM job postings published between January and September and used the company's job postings API together with the OpenAI API to analyse the listings. The analysis found that GTM engineering postings were up 205% from 2024. It reported an average advertised salary of $127,500, while SQL and Python each appeared in 38% of postings.

The same analysis found that nine out of ten responsibilities listed in GTM engineer roles also appeared in RevOps engineer postings. Based on this overlap, the author concluded that GTM engineering and RevOps roles were essentially the same. That is an interpretation of the job-posting data rather than a formal definition of the two occupations.

The Hacker News data provides another way to view the role's visibility among software engineers. A search of Hacker News found seven stories mentioning "GTM engineer", with the highest-scoring story reaching four points. GitHub also shows 107 public repositories carrying the gtm-engineering topic. These figures do not measure employment demand, but they suggest that the title has not yet become a major topic within the broader engineering community.

Salary data provides another useful comparison. The US Bureau of Labor Statistics reported a mean annual wage of $130,140 for sales engineers in May 2025. The mean annual wage for market research analysts and marketing specialists was $89,490. Bloomberry's advertised average of $127,500 for GTM engineers is close to the sales engineer figure and substantially above the marketing specialist figure. These are different occupations, so the figures should be treated as comparisons rather than direct evidence that GTM engineers are paid at a particular level.

There is also a frequently cited figure claiming that 54% of fast-growing B2B SaaS companies employ at least one GTM engineer. The Signal reported this result after examining 63 B2B SaaS companies that it tracks. This is a relatively small, selected group and should not be treated as representative of the wider B2B SaaS market. The same article cites a 3% to 7% figure for private B2B SaaS companies generally, but the source linked for that figure is an AI-generated chat artifact rather than an underlying dataset. That figure therefore should not be presented as an established fact.

The available evidence supports a more measured conclusion. Employers are increasingly looking for people who can build and manage revenue systems using code, data, automation, and AI. The work itself is becoming more important as go-to-market teams rely more heavily on technical systems.

The title is less important than the responsibilities behind it. Some companies may call this role GTM engineering. Others may use titles such as RevOps engineer or growth engineer. What matters is whether the person can build, maintain, and improve the revenue systems described in the previous section.

For more background, see what a GTM engineer actually does, which covers the role's definition and basic hiring considerations.

What should you hire for?

When hiring a GTM engineer, focus first on their ability to make good decisions about systems. Understanding sending infrastructure comes next. Familiarity with specific tools should come last because platforms change frequently.

SQL and Python each appeared in 38% of the GTM engineer job postings analysed by Bloomberry. This suggests that technical skills are already a common part of the role. However, knowing a programming language or a particular platform is not enough. The person also needs to understand how the systems they build can fail.

A few interview questions can also help distinguish someone who builds systems from someone who mainly operates existing ones.

Ask what would happen if a data provider changed the format of its response. A strong systems-focused answer should cover validation, alerts, and fallback processes. Simply checking a dashboard would not address the underlying problem.

Ask how they would identify whether a qualification model had started misclassifying a particular segment. Look for an approach based on sampling and measurement. Waiting until reply rates fall would mean discovering the problem after it has already affected the pipeline.

You can also ask the candidate to explain how they would set up a new sending domain from scratch. Their answer should cover authentication, including DMARC and alignment, as well as unsubscribe handling and spam-rate monitoring. If these areas only come up after prompting, they may have limited experience managing the underlying infrastructure themselves.

Do not place too much weight on the specific platforms a candidate has used. The tools supporting GTM systems will continue to change, particularly as AI takes over more routine work. The more durable skill is understanding how the system works and how to keep it working when individual tools change.

AI fluency should also be treated as a baseline rather than the main differentiator. As discussed in AI fluency as a professional baseline, AI skills are becoming increasingly common across technical roles. A GTM engineer should know how to use AI effectively, but their ability to design, test, monitor, and improve the systems around it is a separate skill.

What does this mean for the SDR team you already have?

AI is more likely to change the SDR role than eliminate it. As AI makes it easier to produce a competent first-touch message, the difference between an average and a strong SDR becomes less about writing that message. It becomes more about understanding what is happening around it.

SDRs can play an important role in identifying problems within the outbound system. A rep who notices that replies from one segment have suddenly fallen, spots errors in a batch of enriched contacts, or questions a qualification rule that keeps producing poor-fit accounts is providing valuable feedback. These observations can help the technical team identify problems before they affect the wider pipeline.

The practical response is to give SDRs better visibility into the system. They should be able to see relevant pipeline metrics, including bounce and complaint rates rather than focusing only on meetings booked. There should also be a straightforward way for them to report inaccurate data or poor targeting, with someone responsible for investigating those reports.

Volume should also become less important as a measure of success. Sending more messages is not necessarily better when email providers are actively monitoring authentication, spam rates, and sender reputation. The goal should be to generate useful conversations without damaging the infrastructure that makes future outreach possible.

The GTM engineer and SDR team therefore perform different functions. The engineer builds and maintains the system. SDRs use that system to have conversations with prospects and provide feedback about what is working. AI can make the underlying system more powerful, but human judgement remains important on both sides.

Where RocketDevs fits

The skills that make GTM engineering valuable are fundamentally engineering skills applied to revenue operations. They require someone who can understand systems, work with data, write code, and make sound decisions when an automated process does not behave as expected.

That is the type of capability RocketDevs' vetting process is designed to assess. Every developer completes 6–8 hours per developer of vetting, and RocketDevs accepts the top 2% of applicants. The process is designed to identify developers who can build and solve problems rather than simply work with a particular tool.

If you need engineers who can treat your go-to-market stack as software and build the systems around it, build a vetted team with RocketDevs.

For a broader look at how AI is changing technical hiring, read what changes about hiring when every developer uses an AI agent.

Conclusion

AI has changed the economics of outbound sales. Writing a convincing email can now take seconds. Research can be automated. Personalisation can be generated at scale. The cost of producing another message has fallen dramatically.

That does not mean the rest of the outbound system has become simple.

Someone still needs to decide which accounts are worth contacting. The data still needs to be accurate. Qualification models still need to be tested. Sending domains still need to protect their reputation. Gmail and Yahoo still enforce authentication and spam requirements. Unsubscribes still need to work. And when something quietly breaks, someone needs to notice before thousands of prospects are affected.

That is where the GTM engineer comes in.

The role is not valuable because it gives a company another person who can use AI to write sales copy. AI has already made that capability widely available. The value comes from having someone who understands how data, code, automation, AI, CRM systems, and outbound infrastructure fit together.

This is also why the job title itself matters less than the underlying capability. Some companies will call the person a GTM engineer. Others may use RevOps engineer, growth engineer, or another title. The terminology will continue to change as the technology changes. The need for someone who can build and maintain the system does not depend on what that person is called.

The evidence for the emerging role is real, but it needs to be interpreted carefully. GTM engineering job postings have grown rapidly in the available posting data, while much of the work overlaps with established RevOps responsibilities. Salary figures and job-posting growth show that employers are putting money behind the capability, but they do not establish a completely new profession overnight.

What is changing more clearly is the nature of the work. As AI takes over more routine execution, the advantage shifts towards people who can make decisions about the systems surrounding that execution. They need to know when to trust an AI output and when to question it. They need to understand what happens when a data provider fails. They need to recognise when a model starts drifting. They need to know why an email is being rejected and what to change before the problem spreads.

In other words, AI makes the message cheaper. It makes the system more important.

For companies hiring GTM engineers, that distinction should shape the entire hiring process. Do not hire primarily for familiarity with whichever tools happen to be popular today. Look for systems judgement. Test their ability to work with messy data. Test their understanding of deliverability. Test whether they can write the code that connects the pieces. Most importantly, test whether they can identify when an automated system is wrong.

The strongest GTM engineering capability is therefore not about replacing humans with AI. It is about building a revenue system in which AI can do what it does well while people remain responsible for the parts that require judgement.

That is the real shift taking place in outbound. The message is no longer the scarce resource. The system is.

Frequently asked questions

Will AI SDRs replace GTM engineers?

AI SDR tools are designed to automate parts of outbound such as drafting and sending messages. That does not remove the need to build and manage the systems around those activities. Research on AI's "jagged frontier" has found that people can perform worse when using AI on tasks outside its effective capabilities. Email providers also enforce authentication and spam-rate requirements, so companies still need someone who can design, monitor, and maintain the systems that support outbound.

Why are GTM engineers in demand?

Generative AI has made producing sales outreach much cheaper. This has shifted more of the technical work towards the data, qualification, automation, and sending systems that support that outreach. One analysis found that GTM engineering job postings increased by 205% year on year. However, the same analysis found substantial overlap between GTM engineering and RevOps responsibilities, showing that much of the underlying work already existed under other job titles.

What skills does a GTM engineer need in 2026?

A GTM engineer needs strong data judgement and an understanding of deliverability. They also need enough programming knowledge to connect systems and automate workflows. SQL and Python each appeared in 38% of GTM engineer postings in the widely cited Bloomberry analysis. Practical knowledge of SPF, DKIM, DMARC, unsubscribe requirements, and spam-rate monitoring is also important because Gmail and Yahoo enforce requirements around these areas.

Is a GTM engineer worth it for an early-stage startup?

It depends on how important outbound is to the company's growth strategy and how much outbound it sends. A company sending large volumes of email may benefit from having someone who can maintain targeting accuracy and protect its sending infrastructure. A company that uses outbound only occasionally may be able to cover the same responsibilities through a technically capable RevOps or growth hire.

The important question is therefore not whether a startup needs someone with the title "GTM engineer." It is whether someone on the team has the technical skills needed to build, monitor, and improve the systems that support its go-to-market strategy.

James Hitch, COO at RocketDevs.LinkedIn

Sources

• Bloomberry, I analyzed 1,000 GTM engineering jobs, 2025, updated 2026

• Gmail Help, Email sender guidelines

• Gmail Help, Email sender guidelines FAQ

• Yahoo Sender Hub, Best practices

• IETF RFC 8058, Signaling One-Click Functionality for List Email Headers

• Brynjolfsson, Li and Raymond, Generative AI at Work, The Quarterly Journal of Economics

• Dell'Acqua et al., Navigating the Jagged Technological Frontier, Organization Science

• Cui, Demirer, Jaffe, Musolff and Peng, The Effects of Generative AI on High-Skilled Work, Management Science

• JetBrains Research, AI coding agent adoption 2026

• Stack Overflow, Developer Survey 2025: AI

• GitHub API, n8n repository

• GitHub, gtm-engineering topic

• Hacker News via Algolia, "GTM engineer"

• US Bureau of Labor Statistics API, Sales engineers mean annual wage

• US Bureau of Labor Statistics, Market research analysts and marketing specialists

• PyPI Stats, hubspot-api-client

• The Signal, 54% of the fastest-growing B2B SaaS companies have a GTM engineer