How to Fix SaaS Go-to-Market: ICP, AI, Pricing
For SaaS companies, go-to-market failure rarely starts with a bad product. More often, it begins with misdirected attention: chasing the wrong customer, overinvesting in the wrong channels, stapling "AI" onto the roadmap without a real data advantage, or using a pricing model that doesn’t fit cost structure or customer behavior.
That’s the central lesson from a recent conversation with investor and former SaaS operator Prasant Chilikuri of Soul Street Ventures. His perspective is useful because it sits at the intersection of operator reality and investor discipline. He isn’t arguing for growth theater, nor for a purely conservative playbook. Instead, he makes the case for something many scaling SaaS businesses need right now: focused execution tied to real customer value.
For founders and operators of $10M+ B2B, SaaS, and e-commerce companies, this is more than startup advice. It’s a reminder that durable enterprise value comes from getting a few foundational decisions right:
Who exactly you serve
Where those buyers actually pay attention
Whether your AI strategy creates real defensibility
How your pricing model aligns to both usage and margin
Key Takeaways
Tighten your ICP before scaling spend. Revenue from edge cases can distract teams from the core market that can actually support repeatable growth.
Meet buyers where they already are. Default channel choices like LinkedIn or Google Ads are often lazy assumptions, not strategy.
Go slow to go fast. Strong GTM discipline starts with market clarity, not with trade shows, swag, or "spray and pray" ad budgets.
Treat AI as a product and data strategy, not a buzzword. The real value is less in generic copilots and more in proprietary data, prediction, and workflow transformation.
Consumption pricing is likely to expand. As compute costs remain variable, especially for smaller vendors, pricing will need to reflect actual usage more often.
Founders need coaching beyond capital. Board governance, hiring judgment, KPI management, and execution discipline are rarely taught but become critical as companies scale.
Durable businesses win by earning the customer repeatedly. Retention discipline often reveals more about product strength than aggressive top-line growth does.
The Real GTM Problem: Founders Lose Focus Too Early
One of the most important points in the discussion is also one of the least glamorous: many founders get distracted.
Early traction can be deceptive. A company builds a product to solve one painful problem for one specific buyer, then discovers adjacent revenue opportunities. Those opportunities can feel validating. They create motion. They may even help cash flow. But they can also distort the company’s strategic center.
That matters because SaaS scaling depends on repeatability, not just possibility.
When teams stretch too broadly, several things usually happen:
messaging becomes vague
roadmap priorities drift
sales cycles become less predictable
CAC rises because channels and personas multiply
customer success inherits a fragmented client base with uneven needs
In other words, the business starts optimizing for "available revenue" instead of "scalable revenue."
For operators in growth-stage SaaS, this is a familiar trap. A few enterprise exceptions or custom deals can create the illusion of product-market expansion while actually weakening the operating model. Chilikuri’s point is that GTM discipline starts by revisiting the fundamentals:
Questions leadership teams should force themselves to answer
Why was this product built in the first place?
What specific business problem does it solve?
Which customer profile feels that pain acutely enough to buy quickly?
Which segments are strategic, and which are merely opportunistic?
What signals prove that your best-fit customers are receiving repeatable value?
This is basic strategy work, but many teams postpone it once revenue begins to arrive. That delay is costly.
ICP Is Not a Slide. It’s an Operating Constraint.
Many companies say they know their ideal customer profile. Far fewer operate like they do.
In practice, a strong ICP should constrain decisions across:
product roadmap
content strategy
paid acquisition
outbound prospecting
partner strategy
hiring profiles for sales and customer success
The interview highlighted a practical truth: customer location is not just geographic; it’s behavioral. Your ICP "lives" somewhere - professionally, digitally, and informationally. They consume certain media, trust certain communities, and respond to certain proof points.
That sounds obvious, yet GTM plans often begin with channel defaults rather than buyer reality.
For example, many B2B teams start with LinkedIn because it is the expected move. But if your buyers are in plant operations, field service, logistics, or niche vertical workflows, LinkedIn may be far less effective than industry communities, trade ecosystems, job boards, trade publications, referrals, or direct account-based outreach.
The strategic implication is straightforward:
The best channel is not the most modern one. It’s the one your buyer already pays attention to.
For larger SaaS businesses, this is especially relevant when entering new verticals. Too many expansion efforts assume that proven channels in one segment will translate directly into another. They often don’t.
Why "Efficiency" in GTM Still Gets Misunderstood
Go-to-market efficiency is frequently discussed as a budgeting issue. But in practice, it is first a decision-quality issue.
The conversation criticized wasteful spending patterns - especially broad, low-conviction channel activity done for the sake of visibility. That includes spending heavily on content built mainly to satisfy search algorithms, or distributing budget across too many unproven tactics without enough strategic coherence.
This is a useful distinction for executive teams. Waste in GTM usually comes from one of four sources:
1. Channel mismatch
You are present where it is easy to buy impressions, not where your buyers make decisions.
2. Weak value communication
Even when the right audience sees you, the message doesn’t clarify the business outcome clearly enough.
3. Lack of testing discipline
Campaigns run too long without meaningful learning loops, or they are changed too quickly to produce signal.
4. Tactical activity without strategic sequencing
Teams spend on brand, paid media, events, and outbound simultaneously before establishing which motion converts best.
Chilikuri’s broader point is that efficiency is not austerity. It is intentional allocation.
That aligns closely with how mature operators should think about growth investments. The question is not "How do we spend less?" It is "How do we spend where conviction is highest and learning is fastest?"
Traditional Marketing Isn’t Dead. Undifferentiated Marketing Is.
One of the most useful ideas in the discussion is that so-called traditional methods still work when they are applied creatively and precisely.
That’s an important correction to the modern SaaS reflex of assuming every scalable GTM system must begin with paid search, polished content engines, and social distribution. In reality, many markets are still moved by specificity, relevance, and surprise.
The point is not that every company should use direct mail or physical promotions. It’s that founders should avoid confusing "common" with "effective." If everyone in your category runs the same playbook, differentiation often comes from using a channel - or a tactic within a channel - with sharper contextual fit.
For leadership teams, the takeaway is to evaluate channels by:
buyer concentration
urgency of the problem
cost to create attention
trust transfer
sales cycle complexity
Not by trendiness.
AI Strategy: Stop Asking Whether to Add AI and Start Asking Where the Defensibility Is
The discussion treats AI as a structural shift, not a passing feature wave. That is the right framing. But it also warns against superficial adoption.
Many founders now feel pressure to present an AI story because the market expects one. Investors ask about it. Buyers ask about it. Competitors advertise it. As a result, companies often bolt AI onto positioning before they have answered a more important question:
What exactly does AI make possible in this business that was not previously feasible, scalable, or economical?
That’s the difference between an AI label and an AI strategy.
A weak AI strategy usually looks like this:
generic assistant features
minor productivity enhancements
shallow automation with little differentiation
dependence on the same foundation models everyone else can access
A stronger AI strategy usually includes one or more of the following:
proprietary or hard-to-replicate data
embedded workflow context
predictive outputs that improve decisions
compounding performance from usage over time
measurable operational or financial outcomes
This is where the conversation becomes especially useful for serious operators. The defensibility in AI is often not the model itself. It is the combination of:
unique data access
domain-specific workflows
trusted user behavior
system-level integration
outcome accountability
That means the question for SaaS leaders should not be "Should we build a copilot?" but rather:
What proprietary signals do we already possess?
What decisions in our customer workflow are still under-optimized?
Where could prediction, summarization, anomaly detection, or recommendation create hard-dollar value?
What insight could we surface that a customer cannot easily produce on their own?
Proprietary Data Is Becoming the New Strategic Moat
A major thread in the discussion is the growing value of proprietary data. This deserves more emphasis because it may be the most consequential point for enterprise SaaS economics over the next several years.
Commodity software features can be copied. Commodity AI features can be copied even faster.
What is harder to copy is a system that has:
accumulated workflow data across a specific domain
structured that data cleanly over time
connected it to customer outcomes
trained internal logic around decisions or predictions
For SaaS businesses already at scale, this creates a strategic opportunity hiding in plain sight.
Many companies are sitting on underutilized data assets inside:
customer communications
support tickets
implementation workflows
usage logs
financial events
operational throughput
field service records
transaction or marketplace behavior
The opportunity is not just to "analyze" that data more elegantly. It is to turn data exhaust into decision support, automation, forecasting, or risk reduction.
This matters because buyers increasingly pay premium pricing for software that does one of three things:
Saves labor
Improves decision quality
Creates revenue or margin visibility
AI can enhance all three - but only when it is fed with context-rich data the system can uniquely interpret.
Why AI Adoption Still Feels Early, Even If the Hype Cycle Feels Mature
Another useful nuance from the discussion: broad market awareness of AI has accelerated faster than practical trust in AI.
That’s a gap enterprise teams should pay attention to.
Yes, executives now understand that AI is not going away. But many still do not trust it enough to hand over sensitive workflows without guardrails. The reasons are rational:
output inconsistency
unclear explainability
governance concerns
privacy and security issues
weak integration into existing systems
uncertain ROI relative to human process
That means there is still a large opening for SaaS companies that can productize trust, not just intelligence.
In B2B software, trust comes from:
clear auditability
constrained use cases
visible accuracy improvements
human-in-the-loop workflows
strong data governance
integration with existing systems of record
This is why many enterprise AI wins are likely to come not from broad "do everything" interfaces, but from targeted applications with specific operational accountability.
Pricing Is Shifting Because Cost Structure Is Shifting
The discussion’s comments on consumption-based pricing are especially relevant for SaaS leaders revisiting monetization.
The old seat-based model works well when value scales roughly with human users. But that logic weakens when:
agents perform work on behalf of users
compute cost varies significantly by usage
some customers generate far more infrastructure load than others
product value is linked to outputs, not logins
This creates pressure toward more consumption-oriented pricing, especially for AI-enabled products.
Why consumption pricing is gaining traction
It aligns revenue more closely with underlying compute cost
It can feel fairer to low-usage customers
It supports expansion revenue from heavy users
It better matches the economics of agentic workflows
Still, it is not frictionless.
Customers generally prefer pricing they can forecast. Finance teams like predictability. Procurement teams like comparability. Seat pricing is easy to understand, benchmark, and budget.
Consumption pricing introduces anxiety unless it is packaged well.
To make consumption pricing work, vendors need to reduce uncertainty
That usually means:
clear usage definitions
spend controls or caps
transparent metering
predictable unit economics
strong onboarding and reporting
packaging that links usage to value
For example, if a customer understands that each unit of usage corresponds to a meaningful business action - an automated response, a processed document, a resolved workflow - they are more likely to view usage-based pricing as rational rather than risky.
For scaling SaaS companies, the deeper lesson is this:
Pricing model design is now intertwined with infrastructure economics.
This is especially true for smaller and mid-sized vendors building on third-party cloud and model providers. If backend cost scales with customer activity, fixed pricing may compress margins faster than expected.
What Founders Often Need Most Isn’t Capital. It’s Operating Maturity.
One of the stronger themes in the conversation is that many founders are excellent at identifying a problem but underprepared for the operating demands that follow institutional growth.
That gap becomes visible in:
board communication
KPI management
executive hiring
organizational design
accountability systems
difficult people decisions
This is not a criticism. It is simply the reality that building a company and running a scaled company are different skills.
For executive teams beyond the earliest stage, this matters because GTM breakdowns often trace back to leadership operating gaps, not just market conditions. A weak ICP, scattered channel strategy, or inconsistent pricing approach is frequently a symptom of something deeper:
unclear decision rights
underdeveloped planning cadence
lack of analytical rigor
poor truth-telling inside the leadership team
One revealing point from the discussion was that some founders tell investors what they think investors want to hear instead of what is actually happening. In practice, that behavior damages the company more than it protects it. If leadership teams can’t surface reality quickly, they can’t correct it quickly.
For operators, that means the real advantage is not optimism. It is fast truth recognition.
A Better GTM Standard for SaaS Leaders
Taken together, the conversation points toward a better standard for SaaS go-to-market - one that is especially relevant for companies trying to scale efficiently in a post-zero-interest, AI-disrupted environment.
That standard looks like this:
1. Start with strategic precision
Know exactly who the product is for and what pain justifies the purchase.
2. Build channel strategy from buyer behavior
Use evidence, not convention, to choose where to invest.
3. Treat AI as an economic lever
Focus on use cases that create new insight, labor leverage, or differentiated outcomes.
4. Align pricing with both value and cost
Your monetization model should make sense for customers and preserve margin under real usage conditions.
5. Develop leadership maturity alongside product maturity
Execution quality often determines whether good products become great businesses.
Conclusion: Build a Business That Earns Its Keep
The most durable idea in the discussion is also the simplest: build a business that continues to earn the customer.
That mindset has consequences. It pushes companies to stay close to value delivery, remain honest about retention, sharpen their ICP, and avoid growth tactics that impress the market but weaken the business. It also creates a healthier lens for evaluating AI and pricing innovation. Not "Is this exciting?" but "Does this create durable customer value and sustainable economics?"
For founders and senior operators, that is the real fix for SaaS go-to-market.
Not more motion. More clarity.
Not broader reach. Better fit.
Not AI for optics. AI for advantage.
And not pricing innovation for novelty’s sake, but for a business model that can scale without breaking.
Source: "Ep. 182 - What Most SaaS Founders Get Wrong About Go-to-Market" - The SaaS Backwards Podcast, YouTube, Dec 5, 2025 - https://www.youtube.com/watch?v=Dn-8vkgEtWc