Your advertising technique could also be caught in 2025.
When you’re rigorously crafting segment-based campaigns, your rivals could be deploying AI brokers that know precisely when your prospects are prepared to purchase — generally even earlier than they do.
“Advertising primarily based on segments is turning into outdated,” says Alex Lindahl, go-to-market (GTM) engineer at Clay. On this unique dialog, he challenges typical advertising knowledge and introduces a revolutionary method that is producing 45% reply charges on chilly outreach. He additionally reveals how AI orchestration is reworking buyer engagement from a scheduled broadcast into an clever dialog.
As somebody deeply concerned in advancing these advertising frontiers, Alex can also be set to share his insights and hands-on methods throughout an upcoming workshop on the AI in Motion roadshow in Atlanta in June 2025.
This is not simply one other AI suppose piece — it is a sensible roadmap for creating what Clay calls “Go-to-Market Alpha,” the aggressive edge that comes from continuous experimentation and adaptation.
This interview is a part of G2’s Q&A sequence. For extra content material like this, subscribe to G2 Tea, a e-newsletter with SaaS-y information and leisure.
To observe the complete interview, try the video beneath:
Contained in the business with Alex Lindahl
Many entrepreneurs are drowning in knowledge however nonetheless combating actionable insights. In your expertise working with Clay’s shoppers, how does an AI-first method basically change the best way entrepreneurs ought to take into consideration their buyer knowledge?
When enthusiastic about knowledge utilization, it is vital to contemplate the various kinds of knowledge which might be on the market. Now we have structured knowledge, present in databases, CRMs, and product utilization metrics. Then there’s unstructured knowledge scattered throughout the online within the type of weblog posts, job postings, monetary reviews, and numerous different sources. So that you must acknowledge that every firm has a singular preferrred buyer profile (ICP) and distinct personas that require particular knowledge units. So the bottom line is determining the place you possibly can supply that knowledge, each internally and externally.
With the appearance of AI, we will now deploy brokers to assemble unstructured knowledge from across the internet, analyze it successfully, extract significant insights, and develop actionable suggestions. The true energy comes from bringing these two parts collectively and mixing various knowledge sources with AI capabilities to speed up understanding and resolution making.
Traditionally, knowledge utilization has largely been handbook, typically restricted to primary segmentation for strategic choices. However immediately, we will apply AI throughout totally different components of the GTM funnel and enterprise processes to dramatically enhance how we collect data, analyze patterns, and take motion primarily based on complete knowledge insights.
Provided that the hole between producing insights and really implementing them is a continuing problem in advertising, what frameworks do you advocate for organizations to show AI-generated insights into concrete enterprise actions systematically? Are there particular instruments or processes that you’ve got seen work significantly properly?
Yeah completely. We consider Clay as a system of motion. So inside Clay, you possibly can have totally different columns powered by AI or knowledge sources, and AI can analyze and construction this knowledge successfully. As an example, if you happen to’re concentrating on a VP of Advertising, AI can align the correct case research or weblog put up particularly for that particular person. The secret’s making use of AI in a structured manner throughout totally different workflow sorts.
AI brokers additionally come into play right here. Traditionally, ICP personas, and advertising and gross sales performs have existed individually with no cohesive timing component connecting them. What I am now calling “preferrred buyer alerts” defines particular indicators that reveal potential enterprise alternatives or challenges the place your product can present worth. This method strikes us away from conventional segments towards signal-based engagement.
And so I got here up with this framework referred to as “DRIVE” that constructions this method.
D – Detect: Establish the sign and decide learn how to detect it and the way ceaselessly to seek for it.
R – Evaluate: Analyze what the sign implies. You may leverage AI right here — for instance, if an organization is hiring DevOps engineers, AI can assist conclude that they are scaling their engineering staff. This overview helps decide whether or not the sign warrants motion.
I – Establish: Uncover challenges the corporate could face in reaching its objectives. AI excels right here when educated by yourself knowledge, weblog posts, case research, and their outcomes.
V – Worth: Align your resolution’s worth to the particular scenario the corporate is experiencing.
E – Execute: Take applicable motion primarily based on the sign.
The perfect implementation entails always-on alerts consistently monitoring outlined account lists. This lets you goal particular accounts exactly when related alerts emerge, indicating well timed alternatives for engagement.
Remodel your AI advertising technique.
Be part of business leaders at G2’s free AI in Motion Roadshow for actionable insights and confirmed methods to reimagine your funnel. Register now
Many entrepreneurs are experiencing “immediate fatigue” from frequently refining their AI directions. What frameworks or approaches do you advocate for creating extra sustainable, reusable immediate methods?
I personally discover immediate engineering extremely enjoyable and highly effective. Particularly as a non-technical particular person, as a result of now principally everybody’s turning into an engineer. You need to use platforms like Lovable to even immediate your option to creating an software. However within the GTM world, there are tons of prompts that you would be able to create, tons of prompts that you would be able to motion.
So, how do you go from simply taking a immediate and refining it over time?
There are a number of approaches to create repeatability along with your prompts. One methodology is creating a customized GPT by embedding your refined immediate right into a specialised mannequin, enriching it with contextual supplies and studying assets. This creates a reusable device that constantly generates advertising belongings or supplies aggressive evaluation with out requiring you to recreate the immediate every time.
One other implementation technique entails AI brokers. At Clay, we make the most of an AI agent that executes customized prompts for analysis functions. As an example, you would possibly direct the agent to find out whether or not a selected firm has a purchasing cart on its web site. The agent will go to the location, analyze it, and return with a definitive reply or related rationalization.
The true energy comes from making use of this functionality throughout a number of accounts — whether or not lots of, hundreds, and even lots of of hundreds — permitting you to scale your immediate’s performance effectively.
The third method integrates these prompts straight into operational workflows. Take into account a easy instance: When an inbound lead arrives, you possibly can apply a immediate to categorize the particular person’s title primarily based on seniority or job operate. A subsequent layer would possibly then advocate essentially the most applicable case research to ship them. This creates a sequence the place AI interventions result in progressively higher knowledge or motion outputs.
The method requires some preliminary funding in creating, refining, and testing your prompts. Nonetheless, as soon as optimized, you possibly can deploy them by customized GPTs, AI brokers, or workflow integrations the place they’re going to operate repeatedly inside their designated context. This eliminates the tedious cycle of repeatedly copying and pasting prompts into ChatGPT, which rapidly turns into exhausting.
Your “Construct the Prompts & Motion the Insights” session within the “AI in Motion Roadshow” guarantees to assist entrepreneurs rethink their present method. What transformation do you sometimes see in entrepreneurs’ considering from the start to the tip of this workshop?
I believe what lots of people are realizing is that advertising primarily based on segments is turning into outdated. The brand new method is utilizing alerts to align the correct advertising or gross sales play to your ICP or persona. So, we’re seeing a shift in considering — alerts at the moment are taking priority over segments.
Reasonably than creating typical campaigns the place you would possibly ship one message this week and a distinct one two weeks later with maybe just a few variations, we’re now focusing extra on listening to the market, accounts, or people. This enables us to use actions or ship advertising messages exactly once they’re most certainly to resonate with the recipient. That is one of many key insights folks sometimes take away from our workshop.
One other realization is that immediate engineering, whereas sounding advanced or intimidating, is definitely fairly accessible. You solely want to grasp some fundamentals to enhance your prompts over time. You may start with a primary immediate and easily ask AI learn how to improve it or what parts could be lacking. This back-and-forth dialogue with the AI helps refine your immediate primarily based on its suggestions, and you’ll incorporate these options and even have the AI rewrite the immediate solely for you.
You simply have to ask the correct questions and take into consideration what the output is that you really want, so you possibly can articulate that clearly to AI. And maybe essentially the most vital takeaway is discovering how simply you possibly can create a customized GPT, apply your refined immediate, and use it constantly to make that immediate repeatable.
This additionally accompanies the broader notion that everybody is basically turning into an engineer immediately. You may create functions, construct merchandise, present AI directions, and develop AI brokers — capabilities that weren’t beforehand accessible to non-technical folks.
It is thrilling that people with out technical backgrounds can now carry out extra technical features by leveraging AI, making all the course of not simply productive however pleasurable as properly.
Alex Lindahl
GTM Engineer at Clay
Your workshop covers “AI orchestrations.” Might you clarify how this idea particularly applies to advertising workflows and supply an instance of how this has improved marketing campaign efficiency?
So, quite a lot of the orchestrating of AI brokers will be achieved by a number of approaches. In Clay, it is fairly simple — you arrange a column that applies your AI agent directions repeatedly throughout an account checklist. Alternatively, you possibly can deploy an ambient agent that displays for particular alerts or implement a chat agent that engages with prospects to qualify leads.
Considered one of my favourite examples entails configuring a customized sign to trace sure key phrases on LinkedIn. When these key phrases seem in posts, each the content material and poster are robotically imported into Clay. The system then makes use of AI to research the particular person’s title, persona, and message content material. Based mostly on this evaluation — significantly specializing in persona and title — it determines whether or not to proceed with message creation or take no motion.
The subsequent section entails AI analyzing the put up and finishing a easy template like, “I simply noticed your put up on [blank].” with AI filling in that clean house. This creates a customized snippet that you would be able to incorporate into your outreach. From there, an automation can set off both a connection request, a LinkedIn message, or an e mail to that particular person.
For this specific marketing campaign, I truly noticed a outstanding 45% reply charge, which is completely insane by business requirements. This success stems from incorporating timing parts and just-in-time messaging through the use of the detected sign to find out the optimum second for outreach. After all, we’re leveraging AI to filter prospects and guarantee we’re connecting with the suitable seniority ranges.
As you possibly can see, this method considerably boosts the general marketing campaign effectiveness as a result of it is extra customized, well timed, and basically extra related to what that particular person is considering or experiencing in that particular second.
For individuals who acknowledge the potential of an AI-first method however work in conventional organizations, what sensible recommendation would you give to somebody who desires to champion these strategies inside their staff however faces institutional inertia or skepticism from management?
I’d say first develop some degree of consolation and information round AI itself. Take a process you consider could possibly be automated and experiment with instruments like ChatGPT or a customized GPT to streamline it. This preliminary step focuses on bettering your personal productiveness whereas gaining useful prompting expertise and understanding what’s doable and what is not.
By this course of, you may naturally develop into extra environment friendly and develop shortcuts for particular duties, reminiscent of changing name transcripts into follow-up emails. This hands-on expertise builds your basis in immediate engineering.
Subsequent, establish a undertaking inside your staff the place you possibly can create a immediate that is going to assist speed up that undertaking and transfer it ahead. And thru that, you are going to begin to expose your teammates to prompting and AI, and so they’ll most likely be fairly shocked. As soon as you’ve got established this foothold, decide learn how to combine AI into your staff’s common operations — their each day or weekly workflows. This gradual growth demonstrates worth at more and more broader ranges: first private productiveness, then undertaking acceleration, and at last, team-wide implementation.
These progressive successes present the proof wanted to construct a compelling enterprise case. Now you can make knowledgeable suggestions primarily based on demonstrated outcomes: “We have efficiently utilized this at particular person, undertaking, and staff ranges. Let’s scale this method to different departments to realize comparable return-on-investment (ROI), productiveness positive factors, and efficiencies.”
If readers might take away only one perception from this text, what would you need them to recollect?
I’d say crucial takeaways are actually two issues. First, have enjoyable with AI as a result of AI exploration will be genuinely pleasurable. Second, experiment with your personal use instances and inside your organization’s GTM technique.
At Clay, we have developed this idea of “Go-to-Market Alpha,” which is basically discovering methods to do issues in another way to achieve aggressive benefits. You may’t uncover these benefits with out experimenting. Some experiments will not present ROI, however by this course of, you may uncover approaches that work remarkably properly.
The previous playbooks aren’t working as a result of we have stopped experimenting. We have to frequently experiment to search out that new playbook, that new wedge, that new ‘Go-to-Market Alpha’.
Alex Lindahl
GTM Engineer at Clay
So preserve experimenting each personally and along with your staff and firm. You may begin to uncover totally different pockets of “alpha” — aggressive benefits that profit each you and your group.
For those who loved this insightful dialog, subscribe to G2 Tea for the most recent tech and advertising thought management.
Observe Alex Lindahl on LinkedIn to study extra about sign primarily based advertising.
Edited by Supanna Das
