Subtraction as Strategy

A client needed to automate one thing. A sales training funnel. Somebody buys a course, they get enrolled in a membership site, the CRM updates, the sequence starts. Straightforward. I started with GoHighLevel because that's what the client was already paying for and I figured we could keep it in one system.

GoHighLevel couldn't handle the membership provisioning the way we needed it. So we added Make.com as a middleware layer. Make.com worked for about two days until we hit a Zapier dead end on the membership side, because WishList Member's REST API turned out to be broken in a way that wasn't documented, and the Digest authentication protocol it used was incompatible with every automation tool we'd tried. We burned maybe a week. Three tools deep. Each one added to solve the problem the previous one created.

The fix was backing all the way out and finding the one integration path that didn't require an additional layer. One connection, direct, no middleware. The client's funnel works now. It runs on fewer tools than we started with, which is the part that still feels counterintuitive to most people when I describe it. We tried adding our way to a solution and the solution was subtracting back to something that worked.

I keep having versions of this experience. At work, at clients, in my own business. Something isn't working, the instinct is to add a tool, the tool creates a new problem, you add another tool to solve that problem, and now you have three problems and two subscriptions you didn't have last month. The default motion in sales ops is addition. It's so deeply embedded that most teams don't even recognize it as a choice. Pipeline is down? Buy a sequencer. Data is dirty? Layer an enrichment provider on top. Reps aren't hitting quota? Add a conversation intelligence platform so you can listen to their calls. Every solution is an addition. Every conference talk is a product launch. Every vendor's pitch is about what their tool does, never about what you should stop doing.

01Fourteen tools for five people

I manage the tool stack for about 20 reps at my day job. Pipedrive talks to a call scoring platform I built, which connects to JustCall for telephony, AssemblyAI for transcription, Claude for analysis, Google Sheets for some reporting, and a custom payment matching system that listens for Stripe webhooks. Every one of those connections is a potential failure point. Every integration requires monitoring. When something breaks at midnight it's usually a connection between two tools, not either tool itself. I maintain this stack because each piece earns its place and I've verified that. But I also know, from sitting inside it every day, that the cost of each additional tool is not the subscription price. The subscription price is the smallest cost. The real cost is the stitching, the monitoring, the admin time, the context switching, and the data integrity risk every time information moves from one system to another.

A typical 5-person sales team that I've audited is running something like 14 tools. A CRM, two or three enrichment providers because somebody signed up for ZoomInfo and then somebody else signed up for Apollo and nobody canceled the first one, a standalone sequencer even though the CRM has sequences built in, a dialer, a scheduling tool, a conversation intelligence platform, an analytics dashboard layered on top of the CRM's own reporting, a proposal tool, project management, and usually some kind of intent data provider. The monthly cost runs about $4,200. The team uses maybe 35% of the features across all of them. Forty percent of those licenses haven't been actively used in 90 days, which is an industry number that matches what I see in practice almost exactly.

The subscription price is the smallest cost. The real cost is the stitching, the monitoring, the admin time, the context switching, and the data integrity risk every time information moves from one system to another.

Consolidated, that same team needs maybe 5 or 6 tools. A CRM that handles sequences natively. One enrichment source. One signal source. Communication tools. And maybe conversation intelligence if the deal complexity warrants it. Everything else goes. The cost drops to about $1,400 a month. That's $33,600 a year back, which for a 5-person team is roughly a quarter of somebody's salary redirected from software licenses to actual pipeline work.

02Nobody noticed for two weeks

The math is never the hard part. The math is obvious every single time. The hard part is that nobody wants to be the person who turns something off.

I've watched this play out repeatedly. You walk into a team and you show them that they're paying for two enrichment tools that pull from the same underlying data sources, and the reaction is never "great, let's cancel one." The reaction is some version of fear. What if the one we cancel was doing something we don't know about. What if there's a workflow running in the background that depends on it. What if we lose data. What if we need it next quarter for the campaign that marketing is planning. The tool has existed on the credit card for 14 months and nobody can describe what it does, and they're still scared to turn it off.

The tool has existed on the credit card for 14 months and nobody can describe what it does, and they're still scared to turn it off.

I'll tell you what actually happens when you turn it off. Nothing. Nothing happens. Nobody notices for two weeks. Then somebody in finance notices the line item is gone and asks when it was canceled and whether anyone approved it. That's the entire disruption.

The deeper fear, the one people don't say out loud, is that removing a tool feels like going backward. The entire culture of B2B sales operations is oriented around building, adding, scaling, growing the stack. Having fewer tools feels like having less infrastructure, which feels like being less serious, less sophisticated, less ready. A 14-tool stack looks like a company that's invested in its sales process. A 5-tool stack looks like a startup that hasn't gotten around to buying the real stuff yet. The optics are backward and everyone knows the optics are backward and they keep buying tools anyway.

03Four layers and everything else goes

I think about this in terms of what I'd ask about any individual tool if I were auditing it cold. Can you name the last deal it influenced? Not theoretically. Actually name a deal that closed because this tool existed. Does anyone on the team use it at least once a week? Not log into it once a week because they get a notification, actually use it to do something they couldn't do otherwise. Does it duplicate something your CRM already does? That last one catches people constantly. Most modern CRMs have sequences, scheduling, basic enrichment, and reporting built in. Teams pay for standalone versions of features their CRM already includes because they bought the standalone tool in 2022 when the CRM didn't have the feature yet, and nobody went back to check after the CRM added it.

I ran into a stat recently that 65% of SaaS apps in an average org are adopted without IT approval. Someone on the team signs up for a free trial, enters a company credit card, and starts using it. Six months later they've moved on to a different role or a different company and the tool keeps charging. Nobody owns it. Nobody maintains it. It just exists on a credit card statement that somebody in accounting reconciles once a quarter without asking whether each line item still serves a purpose. When the person who bought the tool leaves, the tool becomes an orphan, and orphan tools don't get questioned because nobody remembers who's responsible for them.

When the person who bought the tool leaves, the tool becomes an orphan, and orphan tools don't get questioned because nobody remembers who's responsible for them.

The framework I've landed on for small teams, and I mean teams of 2 to maybe 10 people, is four layers. You need a CRM. That's your single source of truth. Every contact, company, deal, and activity lives there. If a tool doesn't write back to the CRM, it functionally doesn't exist because the data it generates will never inform a decision. You need one data source for enrichment and prospecting. One, not three. Pick the one that covers your ICP best and go deep with it. You need one execution layer for sequencing, email, and task management. If your reps are running sequences in one tool and logging calls in another and managing tasks in a third, you've already lost hours of selling time to tab-switching before anyone picks up the phone. And you need one intelligence source, whatever signal matters most for your specific market. Hiring data, intent signals, technographics, whatever. Pick one. Go deep.

Everything beyond those four layers needs to earn its place every quarter. Not every year. Every quarter. Because the stack has a natural tendency to grow and nobody is structurally incentivized to shrink it.

04Nobody's job description includes remove tools

That's the part that makes me the most tired, honestly. The incentive structure. No vendor will ever tell you to use fewer tools. It's not because they're dishonest. It's because their business model is subscriptions. Their job is to make you believe you need their product. Their sales team is compensated on new logos and expansion revenue. The entire apparatus of B2B software sales is designed to make your stack bigger. The blog posts, the case studies, the ROI calculators, the "see how Company X increased pipeline by 40%," all of it points in one direction: add this tool.

The only people who benefit from subtraction are the operators inside the company and the company itself. The ops person who maintains the stack. The finance person who pays for it. The reps who have to use it. And the outside consultants whose business model doesn't depend on you subscribing to something, which is a small group because most consultants are also resellers or implementation partners for the tools they recommend.

The one-file version isn't elegant. It's not pretty. It's also been running for months without breaking because there's nothing to break.

I built a prospecting system for my own business that runs as a single markdown file. You drop it into an AI session. It runs. That's the whole product. Competitors are packaging similar capabilities into multi-tool agent suites with dashboards and integrations and subscription tiers. I keep looking at those products and thinking about how every additional component is another thing that breaks, another thing that needs updating, another surface area for something to go wrong. The one-file version isn't elegant. It's not pretty. It's also been running for months without breaking because there's nothing to break.

Ninety-four percent of sales orgs say they plan to consolidate their stack. That number comes up in surveys every year. The actual consolidation rate is nowhere close to that. The intention is there. The follow-through almost never happens because the fear is real, the vendor pressure is constant, and nobody's job description includes "remove tools." It's always "evaluate and implement." Never "evaluate and subtract." The language itself is biased toward addition.

I don't think this changes from the vendor side. I think it changes from the operator side, from the person inside the company who actually maintains the stack and knows what gets used and what doesn't. That person usually knows exactly which tools to cut. They've known for months. They just haven't had the cover to do it because removing a tool that a VP approved last year feels like a political move even when the math is obvious.

05Eleven months at a hundred eighty dollars a month

I canceled a tool at work last year that had been on the books for about 11 months. It was an analytics overlay on top of Pipedrive that produced dashboards we could build natively in Pipedrive with about 20 minutes of configuration. I asked three reps if they used it. Two didn't know it existed. The third knew it existed but used Pipedrive's built-in reports instead. The tool cost $180 a month. I canceled it, rebuilt the two dashboards it was theoretically providing in Pipedrive, and nobody mentioned it again. That was 11 months of paying $180 for something three reps didn't know was there.

I suspect most teams have at least two of those. Tools running quietly on a credit card, duplicating something the CRM already does, used by nobody, owned by somebody who left. The first time you go looking for them it's a little shocking how many there are. The second time it's just maintenance.

The hard part is remembering to look.

If your stack is growing but your pipeline isn't: demo.outblox.com