RevOrg leaders are intellectually aware that AI adoption could potentially improve the core revenue metrics. However, they are unsure of how to apply this in practice that produces a demonstrable ROI.
In this article, we examine how a strategy called Shift Left can be successfully used for effective adoption of Agentic AI in RevOrgs.
The core idea of Shift Left is to pull forward latter stage activities to earlier stages of the sales cycle.
The result is compressed sales cycles, better pipeline predictability and revenue acceleration and expansion.
AI and Agentic AI are poised to have a significant impact on revenue organizations. Modern sales leaders are keen to adopt AI, but face major barriers to realizing that vision. They lack a grounded strategy for AI adoption. There is no systemic and standardized usage charter for sellers who are already a bit fearful of their jobs. Without a well defined evaluation framework, leaders don’t know how to filter the noise coming from new products that are launched every day. Above all, they need to temper unrealistic expectations from stakeholders, with little external help, while continuing business as usual.
Let’s take a closer look at the AI adoption strategy, which is one of the core issues sales leaders face. For a more strategic AI adoption, revenue leaders need to embrace a paradigmatic shift and consider a holistic application of AI across the whole sales process. In this article, we examine how one such philosophy from software development can be successfully applied to full sales cycles to bring measurable benefits to RevOrgs.
Shift Left that was introduced to software development around 20 years ago. Conceptually, you can visualize the software development lifecycle as a process that moves from left to right. Shift left moves some of the activities that usually happen late in the process (right), such as testing and security, to earlier in the process (left). This allows issues to be captured earlier which also makes it simpler to fix them.
The process for shipping a product or feature follows a similar linear progression from left to right. You start with gathering market requirements and move on to product specifications, software development and production deployment. Each of these stages are owned by different teams. A company relies on the successful translation of requirements across stages to ensure that the final product reflects the market needs. It is not until the third stage that a working prototype can be developed for even a preliminary market validation. If the intermediate translations fail, there will be a big divergence of the finished product from the market needs resulting in high development sunk costs.
Now imagine a scenario where a sales engineer can accurately capture a customer’s requirements by using AI to generate a prototype, at runtime, in a customer meeting. This shifting left of the prototype development from the third stage to the first stage will significantly make product development predictable and accurate. So while AI is nowhere near replacing the people or the process in product development, it can certainly accelerate time to market, reduce technical debt and improve customer acquisition by enabling the adoption of shift left in the product/feature delivery lifecycle.
Revenue organizations, just like development organizations, are under tremendous fear and confusion as AI becomes more mainstream. Sales leaders are saddled with unreasonable expectations that RevOrgs will undergo an overnight transformation into AI powered beasts that will crush all historical revenue metrics. Without a grounded framework or an internal team to lean on, leaders are tasked to assimilate AI into their business, while conducting business as usual with their sellers who are already grappling with an existential threat that their jobs will be replaced by AI. This could be overwhelming, leading to avoidance instead of embracement.
AI adoption does not have to be intimidating. With a systematic approach, AI can augment, not replace, sellers. Sales cycles can benefit from a strategic and realistic introduction of agentic AI. The challenge is to figure out where and how.
There are many tools and solutions out there that genuinely help sellers in specific activities during the sales process. Sellers adopt them on an ad hoc basis to help them at their work. We won’t discuss those point solutions here. Instead, we will talk about a more strategic perspective that alters how we view the selling process. We will examine how we can borrow the concept of Shift Left from engineering organizations and apply it to RevOrgs for better and more predictable sales outcomes.
A sales cycle can be visualized as a process from left to right. It typically consists of the following stages:
Lead generation: At this stage, someone like a BDR will scan the market to identify companies that fit the ideal customer profile, research the target company to find the people that fit the buyer profile, set up introductory calls to confirm buyer’s intent and hand off a warm lead to the seller.
Qualification and discovery: The seller, sometimes with a sales engineer (SE) in tow, will determine if the prospect has a problem that can be solved and the budget and urgency to buy the solution. This is also a stage where other business and technical aspects are flushed out to ensure that the opportunity is the right one to pursue.
Presentation and demo: Now that the opportunity and offering is aligned, the seller and the SE will do a presentation, often dividing the responsibility between the non-technical and technical parts, including a product demo, between them.
Trial: Before buying a complex technical solution, a buyer will always want to do a proof of concept to ensure that the solution meets all the acceptable criteria, check how it compares with other solutions they are considering and understand how easy it is to integrate with their existing tech stack. The SE, with support from internal teams, runs this stage.
Proposal: Assuming all goes well up until now, it’s time for the seller to formally present a proposal to the buyer. It includes the problem and solution, the ROI and the commercial and contractual terms among others. This is also a stage that involves buy-in from a variety of buyer stakeholders.
Negotiation: This includes a whole gamut of activities such as negotiating price and commercial terms, handling legal, and overcoming objections. New technical needs may creep up here, sometimes because of a genuine reason, while at other times, as a negotiation tactic.
Close: The final step should, in theory, be easy, if everything else was done right until this point. But buyers are humans and thereby, unpredictable. They can stall, get distracted or revisit any of the previous steps. Good sellers are masters of this step and know how to get the signature that turns a prospect into a customer.
Each of those stages has multiple activities. Adoption of a Shift Left paradigm in the above cycle would move many of the later stage activities to the earlier stages in the sales cycle. Agentic AI automations, using an analytical framework, can make that transformation possible.
For example, during the lead generation stage, a BDR can also perform quite a bit of the qualification and discovery, thereby reducing the vetting load on sellers and SEs. As a result, the handoff to a seller will be a highly qualified lead and not just a warm lead. Similarly, AI could ensure that the proposal generation starts right at the discovery phase. As a result, the sales cycle gets transformed from a sequential process to multiple, parallel, overlapping processes.
Introduction of agentic AI in this manner is not intended to replace sellers or redefine the sales cycle. But with a thoughtful adoption, it can compress the sales cycle and increase the probability of favorable outcomes by shifting more activities to the left of the process.
A sales cycle can be neatly represented on paper as a sequential process diagram. However, the reality of an engagement is far more complicated and nuanced. There are multiple activities that occur simultaneously, interleaved, and span many stages. The stages are not as discrete and handoff between stages are not as clean as we would like them to be. For example, customer research is not just limited to lead generation and discovery. It is a continuous process that extends all the way down to the closing stage.
Buyers don’t wait for a specific stage to provide the information that you need. Remember this is a sales process defined by the selling party. Buyers didn’t sign up for that process. They reveal a lot about themselves throughout the journey, not just at discrete points. Depending upon the context of the conversations, they can reveal a lot about the organization, the projects, their fears, their technology and competitors at random points of the sales process.
It is important to capture them early, as soon as they are mentioned. This allows the sellers to be better informed, position themselves more strategically, and drive deeper conversations which builds a trusted relationship with the customers. It also prevents customers from having to repeat themselves, which builds confidence. All things being equal, trust and confidence are often the intangibles that become deciding factors for the buyers.
Buyers engage suppliers over a dozen channels ranging from emails and meetings to Slack messages and industry conferences. Omni channel is a great strategy to stay on top of a buyer’s mind. But it also makes it difficult to collect and synthesize all the buying signals they might be dropping at different places, especially since a seller is not managing all the channels.
Is there a pain point that the technical buyer mentioned in the support ticket during the trial?
Is there something that an executive from the buyer mentioned to your event marketer at a conference that could influence the decision?
Did the buyer ask a question to your website chatbot that might reveal something significant?
Good sellers know this and proactively capture all this information. But it is really hard to do since the process is all manual. Now imagine how hard it will be for them to do it for all their customers. As for the less experienced sellers, this would almost border on impossible.
Agentic AI can be a game changer in this aspect. An AI workflow that constantly collects, collates and assimilates this mountain of information to continuously provide strategic guidance to sellers could accelerate and expand enterprise revenue. Capturing insightful nuggets when customers provide them earlier in the process, rather than later when the seller asks represents a shift left from the buyer’s perspective.
Introducing agentic AI into the sales process can have a material impact on a company’s revenue. It is therefore imperative that there is a way to measure and quantify the impact of AI adoption. But it must be remembered that AI should not be measured in the abstract, on its own. AI is simply the grease that makes the sales machinery move better. As such, the quantifiable benefits to track should still be the standard revenue metrics. What you want to track specifically is if those metrics moved noticeably due to the introduction of AI.
A few key revenue metrics that have the potential to be impacted due to the adoption of Shift Left strategy with agentic AI are as follows:
Sales cycle length
Win rate
Deal size
Seller productivity
Shifting more of the downstream sales activities to upstream can compress sales cycles without redefining it. Sellers can use AI to be constantly up to date on buyers needs and queries. This minimizes lost cycles due to information repetition, further shortening the lifecycle. Shortening sales cycles help pull revenue forward that augurs well for the financial health of the business.
Win rate as measured by the closed won can be expected to improve because of real time awareness of and timely response to buyers needs. Even the “intermediate wins”, as measured by MQLs and SQLs, should be expected to improve due to buyer research and qualification early on in the process.
Deal intelligence can have a significant impact on not just the close rate but also the deal size. Buyers don’t always know, or are not able to articulate their complete needs up front. Sometimes they might think of a need long after the technical discovery call. They might ask about it to your support staff while working through a support ticket. Miss it and you have missed an opportunity to expand the scope of the deal. Sometimes situations change in a company and a newer context might bring up new needs. The buyer, in the middle of the night, might want to ask your chatbot to quickly determine if that need can be met. Miss it and you miss an upsell opportunity. Measuring the increase in deal size since the first forecast can be a great way to measure the impact of AI.
Sellers today spend only 30% or less of their time in selling, the one thing that they are masters of and were hired to do. Seller productivity can increase 3x if you could hand them each an assistant to do their routine work. But that is impractical. However, AI could be that assistant. It can offload all routine work, turn them around in minutes, and in the process freeing up sellers to be hyper productive. Tracking metrics like selling time vs admin time and pipeline/revenue per seller will give you insight on how well AI is performing in this context.
The above is not an exhaustive list of metrics or their usage. The specific metrics that are useful and measurable will vary with industries and companies. But what should remain consistent is the way you approach introducing and measuring the impact of AI:
Do a full audit of your workflows and work with an expert to determine which ones can be implemented with AI. It is important to remember that not all workflows can be automated.
Sequence them using a right priority framework. An often forgotten but useful metric to use in this context is time to value. Prioritize a smaller project with a lower expected ROI if it can be implemented and measured fairly quickly.
Define metrics relevant to that workflow. Don’t invent new metrics. Use the ones that you have always been using.
Get a baseline of those metrics for the past 3 - 6 months.
Apply AI.
Measure again and compare with the baseline.
Adjust or discard based on the data.
It is not easy to capture return on investment. However, if done right, it can provide a powerful narrative of how your core revenue metrics were impacted due to the adoption of agentic AI.
AI can’t replace sellers, at least not yet. But AI certainly has an influential role to play to make revenue orgs more effective. That requires a shift in thinking on how to embrace and introduce AI and temper expectations so that they are grounded in reality. What exactly that looks like will become clearer as the agentic AI landscape becomes more mature. What will undoubtedly be true is that AI literate sellers fueled by Agentic AI enabled workflows will be the force multipliers in enterprise B2B revenue organizations..