AI for GTM: What works now

Big platform shifts lead to massive changes to the way we work and the tools we use. The move from on-premise to cloud sticks out because of Salesforce, who bet on the cloud as the future of software, setting themselves up to take advantage and define a generation of tools. As everyone moved from on-premise software to cloud, a flood of new companies were founded to capitalize on that innovation. Eventually, companies all began building their new tech stacks in the cloud. 

We are in the midst of a similarly colossal platform shift with AI. Incumbents are adding new AI tooling and new entrants are launching every week.The result is a bit of chaos for everyone. It’s not clear which tools you should use for which use cases. Some AI is helpful, some AI is a solution in search of a problem to solve. 

If you’re confused, don’t worry, you’re not alone… so is everyone else trying to decipher the GTM x AI landscape. In this article we’re going to try to demystify the new GTM tech stack and break down where AI is a valuable tool versus a toy. 

AI maturity for GTM 

You’ve probably heard the saying “is it a vitamin or a painkiller?” Well, for AI I’ve started to say, “is it a tool or is it a toy?” We’re in the early innings of the AI revolution and there are some use cases with more maturity than others. Every tool promises outsize results, but it’s still undetermined who can deliver and which use cases are best suited.

The biggest challenges are fourfold:

1. Lack of clarity on use cases - which workflows can AI actually help automate or augment?
2. Lack of clarity on maturity for those use cases - is this still a toy or can it be operationalized?
3. Lack of clarity on tool’s success rate - which products actually do what they promise?
4. Lack of clarity on how to operationalize and scale - for many AI is still being incubated with individuals or small teams, but how do you scale this across an entire GTM org?

We’ll cover the first two in part one of this series, subscribe to our newsletter to follow along. 

Our philosophy on AI

Separating tools from toys begins with an understanding of where AI is useful today in GTM. In our view, there are 4 categories where AI provides outsize impact on GTM use cases. 

Research

Gathering research is limited by the amount of time a human has in a day, access to data/systems, and the ability to bring information together into actionable insights. AI on the other hand, can run through volumes of data and research on the internet in seconds to synthesize results. 

Orchestration

One of the biggest areas of improvement in AI is in the reliable orchestration of workflows powered by AI agents. AI agents can be deployed to automate routine tasks like updating systems with new data, smart routing, or assigning owners. 

Personalization

Recent advances in AI have unlocked a higher bar for personalization. AI can tailor content and write personalized copy that resonates based on research that is performed on the industry, company or person. 

Predictive analytics

Although this is not a new area, AI can analyze vast amounts of data to identify patterns that are helpful in domains like lead scoring. Take sales playbooks where you have a measurable workflow and conversion data - AI can analyze the current playbooks and make suggestions for how to tune and iterate. 

Use cases for AI in Sales  

Instead of going down a market mapping exercise, we think the way to break down where AI is tool vs. toy is to focus on 2 primary sales functions. Unlike the rapidly changing landscape of tools, the jobs to be done in these sales functions will remain relatively unchanged.  

‍Do sales (Reps): The activities reps do to build pipeline and generate revenue. Jobs-to-be-done include:

1. Prospecting
2. Close & expand revenue  
3. Deal admin

Run sales (RevOps and Leadership): The planning, building, measuring and optimizing that goes into the foundation of running a sales organization. Jobs-to-be-done include:

4. Identify & prioritize opportunities
5. Build sales playbooks & enablement 
6. Forecasting 

We’ll dive into each of these jobs-to-be-done below.

Do Sales 

We’ve added a ton of new technology to the “Do Sales” tech stack over the last decade, which rather than increase productivity has led to more time spent on non-revenue generating tasks. Salesforce’s 2024 State of Sales report found that reps spend 70% of their time on non-selling work. 

With AI, some of these tasks can be completely automated, others made faster with an AI-assist, and some are still not a great fit for AI. Let’s look at each use case in the category and where AI fits. 

1/ Prospecting  

Prospecting is the engine that fuels growth for any company. Prospecting is still a very time consuming and disjointed workflow.

The jobs:

• Action prioritized accounts 
• Research and plan 
• Find the right contacts 
• Write compelling messaging (email, calls, social selling) 

Where AI fits in: 

• Action prioritized accounts: Reps don’t have to manually prioritize their focus, AI models tuned by RevOps can score every lead or account. However, be warned that a black box score rarely leads to adoption with reps. We find that surfacing discrete signals (Recently hired VP, hiring, ICP fit) that provide context work better than “Acme corp scored a 75”. 
• AI to surface the right contacts: Similar to above reps also waste a ton of time looking for the right contacts within their accounts. AI can suggest the right buyers by analyzing their ICP fit, previous engagement, or any other signal that might indicate relevance. 
• Research & context gathering: Reps have limited time and ability to go deep on every account to build compelling account plans. With AI, reps can understand the context of an account based on internal and external sources in seconds to inform their outreach strategy.  
• Qualification: AI can assist in summarizing calls and other context to highlight if your reps are properly qualifying opportunities. Follow a specific methodology like SPICED or MEDPIC? Train AI to detect when qualification criteria is met. 
• Email writing assistants: AI is great at giving reps a starting point for emails based on context synthesis on an account, but this content should be up-leveled by a human afterward. 
• Fully automated SDR: Once you have a clear playbook, autonomous prospecting agents or “AutoSDRs” can seem like an obvious fit use case. To date, AutoSDRs are best suited to SMB teams to accelerate experimentation, but AI is not good enough yet for this to be scaled.

2/ Close and expand revenue 

The work isn’t done once you build pipeline. The work required to close revenue is all about knowing your buyer and providing an excellent experience. It’s about being two steps ahead, navigating complex organizations, and building a compelling business case. The sales process is a mix of art and science. 

The jobs:

• Prep for meetings 
• Account planning 
• Multi-threading stakeholders 
• Building business cases 
• Generating quotes for approval
• Meet with and engage champions 
• Run discovery to find new use cases 

Where AI fits in: 

• Account planning: AI can combine 1st party (call notes, recordings, product activity, etc.) with 3rd party signals (10ks, podcasts, news, etc.) to have strong POVs on where the product fits.
• Meeting prep: AI can synthesize context about an account into snackable information that is easy to skim before jumping into a meeting. AI trained on sales methodology can auto-generate questions to ask or insights about customer pain points, budgets, and timelines. 
• Call coaching and insights: AI tools can listen to calls to help identify improvements for the rep and areas of opportunity.
• Stakeholder mapping: AI tools can find the right buyers and analyze their level of engagement in a deal based on available signals. 
• Business case generation: AI can help build data-driven business cases by pulling together relevant data, such as ROI calculations, case studies, and market trends.

“PMM may create templates for sales collateral but often AEs are creating bespoke collateral for their own accounts - AI can be immensely powerful here to speed up that process. Same applies for other content-related tasks like responding to RFPs and personalizing demo environments.” - Mark Goldberger, Head of Enterprise Sales, Ramp

“One of our primary genAI use cases is oriented around delivering customer feedback to our product team. We synthesize customer feedback into insights which are used to improve the product.” - Casey Bertenthal, Head of Sales, Linear

Note all of the above can apply across the funnel from closing business (sales) to expanding and retaining business (post-sales) .

3/ Manage deals

Typically a rep’s least favorite aspect of their workflow is miscellaneous admin tasks that keep the sales process moving. 

The jobs:

• Keeping CRM up to date 
• Managing follow ups after meetings 
• Scheduling meetings 
• Sending collateral 

Where AI fits in: 

• CRM management: AI can automate data entry by capturing information from emails, calls, and meetings, and then updating the CRM accordingly.
• Meeting recap & followup: Use AI to create meeting summaries and create follow up action items. 
• Scheduling assistants: AI scheduling assistants can double as general-purpose AI assistants to track follow-ups and action items from emails and calls. 
• Collateral suggestions: Based on call recordings and meeting notes AI can easily suggest the type of collateral an AE should include in their follow up notes. 

“At OpenAI we take all of our call transcripts and feed them through a GPT in ChatGPT Enterprise, which creates a beautiful recap and follow up email. Super low lift - anyone could set this up in a few minutes and save reps hours per week of follow ups.” - Maggie Hott, GTM Leadership at OpenAI

Run Sales

Leadership and RevOps teams have had a tough couple of years running sales. It’s harder than ever to sell software, teams are forced to do more with less, previously reliable channels no longer work, and every deal is scrutinized. To make matters worse, the ballooning of the tech stack makes rep productivity a challenge.

In 2023, only 28% of reps hit their quota attainment goals. The outlook for 2024 and 2025 are much more positive in part because teams are focusing on sales productivity, reducing tool fatigue, and implementing new AI tools that reduce busywork. Let’s look at each use case in the run sales category and how AI fits in. 

4/ Identify & prioritize fit accounts 

Mapping your TAM, current target accounts, and ICP can be a very manual task. It requires pulling historical data, segmenting that data, running analysis on average contract values, and more. Instead, AI can quickly surface accounts most likely to purchase, identify expansion opportunities, and flag accounts at higher risk for churn. 

The jobs:

• Identify ICP
• Map your TAM 
• Segment into territories & account lists 
• Score accounts 

Where AI fits in:

• Find lookalike accounts: AI can help identify the signals that make accounts a good fit for your business and then use that analysis to find lookalikes to build your target account list. 
• Unique signals: Before AI, companies were confined to their 1st party signals or the handful of 3rd party data providers / intent vendors. With AI, the ability to quickly scrape the internet for very unique signals to your business unlocks even better prioritization and message personalization. 
• Account and lead scoring: AI can develop scoring models that rate potential accounts & leads based on their fit with the ICP. These scores are generated by evaluating factors such as firmographics (e.g., company size, revenue, industry), technographics (e.g., the technology stack used by the company), and intent data (e.g., online behavior indicating purchase intent). AI can analyze historical data to find patterns in your best opportunities and then use those attributes and other hypotheses to predict future opportunities. 

5/ Build sales playbooks 

Playbooks are the lifeblood of sales organizations. RevOps and leaderships use these to train the team on where to focus, what to say, and how to say it. A playbook can be as simple as a lead list and an email sequence or more complex with instructions and enablement for reps. Finding the right playbooks requires experimentation. It’s not always easy to understand what playbooks are working, what playbooks need tuning, or what playbooks you’re missing.

The jobs:

• Analyze closed/won to define ICP (find best fit account attributes)
• Choose signals that correlate to conversion
• Write sales playbooks 
• Enable sales on playbooks 
• Build sales enablement materials (email sequences, messaging, collateral)  

Where AI fits in:  

• Suggested playbooks: AI can analyze past performance of playbooks and recommend new playbooks to try. 
• Build personalized enablement content: AI can learn from existing enablement materials and create infinite personalized variations based on the specific playbook you run, customer segment, industry and more. 
• Rep coaching: AI can make suggestions for how to coach reps to run their playbooks. Maybe reps are not responding fast enough or their messaging on calls isn’t sharp enough. 

“We’re reaching a point with AI now where content production is personalized not just for accounts but for a specific moment in the customer’s journey. Imagine the product team launches a new feature, AI can identify which accounts would find it most useful and generate a custom 1-pager. Or a customer mentions a specific set of problems in a discovery meeting, AI can use that insight to suggest the right collateral to an AE.” - Zeya Yang

6/ Forecasting & planning 

A huge part of RevOps role in running sales is to keep a pulse on the health of the overall business. RevOps looks at the GTM motion as a whole to assess win rates, performance to plan, and forecasting into the future. This is perhaps one of the areas where AI is most mature, tools like Gong and Clari have pioneered how revenue teams get revenue intelligence.  

The jobs:

• Track and monitor pipeline metrics 
• Analyze wins/losses
• Forecast revenue 
• Planning & resource allocation

Where AI fits in: 

• Win/loss analysis: AI can detect patterns in why teams win or loose deals like pricing issues, product fit, or competition.
• Forecasting: AI uses historical data and current performance to predict future revenue attainment. 
• Sales capacity & resource allocation: AI can analyze current and historical performance metrics to help plan optimal resource allocation—such as headcount, budget distribution, and territory assignments

Assess AI maturity and fit

Knowing the use cases is only half the battle. How do you assess whether that AI use case is operationally mature? Is it a ‘Tool’ or a ‘Toy?’

We built a simple scoring system to evaluate the fit of AI for any use case. 

Let’s break it down by criteria:

• Tedious: Is the task repetitive or slow? AI does well with repeating patterns and following simple instructions.
• Time-consuming: Do you spend a significant amount of time on this task each day or week? AI can give you hours back by cutting high-volume work down to a few clicks.
• Data-heavy: Are you trying to synthesize big quantities of data into easy-to-understand bullets or actionable tasks? AI can identify patterns, crunch numbers, and pull out key concepts in seconds.
• Low creativity: Does the task not require critical thinking or thoughtful wordsmithing? AI produces results that read a little robotically, so if your task doesn’t require too much human nuance it’s a good one to automate. 

If it checks all 4 of those boxes it’s a certified AI ‘Tool’ you can deploy. Only checks 1? Maybe still an AI ‘Toy’ and not the right fit today. 

Final thoughts on AI in GTM 

1. AI isn’t going anywhere. It’s already having a massive impact on GTM.
2. The human element of sales can’t be replicated, sales as a profession isn’t going anywhere. While AI has a ton of potential for almost every use case, most use cases are helping augment sales jobs vs. replace jobs.
3. AI autopilot systems like AI SDRs are still in experiment mode and are being used by SMB teams who are resource constrained. 
4. AI is capable of doing a lot of the tedious & manual sales jobs, which will elevate the importance of soft skills for sellers - communication, strategy, relationship building, empathy, and creativity will be highly valuable.

10x your sellers with Pocus AI

Pocus’ bet is that AI will NOT replace sales reps. Instead, AI will eliminate a ton of busy work for the GTM team and level up individual reps to become 10x sellers. 

“Sales has always been both art and science, with AI’s capabilities to automate or eliminate the science, it gives sales teams time to focus on the art. The craft of selling, building relationships, and creativity. Ultimately the result is increasing win rates, accelerating sales cycles, and happier more productive reps.“ - Adam Carr, Head of Global Sales at Miro

How does this bet translate into our current product and roadmap? We’re focused on optimizing and streamlining 3 core workflows: 

1. AI Playbooks - helping reps prioritize and focus their time starts with smart, data-driven playbooks. 
2. AI Research - the top 1% of sellers deeply understand their customers’ problems and build compelling POVs. We’re building tools to make this easy for every rep and scalable across hundreds of accounts. 
3. AI Prospecting - knowing who to reach out to and the best path to break in is a constant battle for reps. We’re helping surface the best contacts to do just that. 

Are you ready to take advantage of AI?

Circling back to the platform change, this graphic from HyperGrowth partners comes to mind. It shows the cycle of change that happens when new technologies emerge. The top 1% of teams are always early to adopt and get all the leverage from these platform shifts. 

The AI platform shift is going to be a huge advantage to teams that embrace it. The AI GTM landscape is still very confusing, so we’re on a mission at Pocus to demystify this new frontier. 

{{widget-community}}

--------

Special thanks to Mark Goldberger, Casey Bertenthal, Maggie Hott, Zeya Yang, Adam Carr, Sam McKenna, Kate Jensen, Andrew Johnston, Brendan Short, and Michael Thompson for contributing valuable insights and sharing feedback during the creation of this AI x GTM breakdown.