The Math
Every department, every dollar

Every Department, Every Dollar

This is the chapter where we stop telling stories and start counting.

Every use case in this book included a savings number. Every department chapter ended with a tally. This chapter pulls all of it together into one place, checks it against the cost of the tools that made it possible, and gives you the net number. The number that matters.

No hand-waving. No "estimated" savings without specifics. Every dollar here traces back to a named SaaS product, a specific annual cost, and a use case that replaces it.

Maya's Marketing Team

Maya's marketing team at a 60-person B2B company ran ten use cases across all six tools.

Marketing's Share of AI Tool Costs

Priya's Sales Team

Priya's sales team at a 40-person SaaS company. Ten use cases.

Sales' Share of AI Tool Costs

James at a 50-Person Company

James at a 50-person company. Ten use cases.

HR's Share of AI Tool Costs

Raj at a 70-Person Company

Raj at a 70-person company. Ten use cases.

IT's Share of AI Tool Costs

Lisa at a 45-Person Company

Lisa at a 45-person company. Ten use cases.

Office Management's Share of AI Tool Costs

Alex, Independent Consultant

Alex, independent consultant with eight regular clients. Ten use cases.

Alex's AI Tool Costs

The Combined Picture

There are two ways to read these numbers. The first is as a company running five departments (Marketing, Sales, HR, IT, Office Management). The second adds the freelancer as a standalone operator. They are different calculations because they share tools differently.

The Company Math (Five Departments)

Now here is where the math gets interesting. Each department chapter attributed AI tool costs conservatively, some allocating the full subscription, others allocating a proportional share. If we were doing separate departmental budgets, those numbers stand. But a company deploying these tools across all five departments shares the subscriptions.

Actual Company AI Tool Cost (Shared)

Read that again. One hundred thirty-eight thousand, seven hundred sixty dollars. Per year. For a mid-size company running these replacements across five departments.

The conservative per-department estimates that attributed full tool costs to each department totaled $132,200 in net savings combined. The real number, where shared subscriptions are counted once, is $138,760. Even the conservative number is significant. The real number is transformative.

The Freelancer Math (Standalone)

Alex's math is simpler. One person, one set of subscriptions, no sharing.

That is $610 per month back in Alex's pocket. Not a corporate budget line item. Real money in a real checking account.

What Goes

Sixty SaaS subscriptions across six chapters. Here they are, all in one place.

That is the long tail of SaaS. Dozens of point solutions, each solving one problem, each charging a monthly fee, each used at a fraction of its capacity. Individually, each one seemed reasonable. Collectively, they are a quiet drain on budgets across every department.

What Stays

This is equally important. Not everything can or should be replaced. Here is what stays and why.

CRM (Salesforce, HubSpot, Pipedrive). Your system of record for customer data, deal tracking, and team coordination. The tools in this book replace the analytics and reporting layers that orbit the CRM. The CRM itself stays.

Email sending infrastructure (Mailchimp, ActiveCampaign basic tier). You need a platform to send emails, manage lists, and handle compliance. The template builders and content creation tiers go. The sending engine stays.

Social media scheduling (basic tier). Scheduling and publishing posts at optimal times is a feature worth keeping. The content creation tier goes.

Call recording (Gong, Chorus core). Recording and transcribing sales calls provides genuine value. The analytics layer that costs extra was replaced. The recording stays.

E-signatures (DocuSign, HelloSign). Legally binding electronic signatures are a compliance necessity. Keep them.

Outbound automation (Outreach, SalesLoft). Communication sequencing with deep CRM integration stays.

Core HRIS (BambooHR base, Rippling base). Employee records, payroll integration, benefits administration, tax compliance. This is not buildable. This stays.

Applicant tracking (Greenhouse, Lever base). The hiring workflow engine stays. Interview scorecards were the only replacement.

Ticketing system (Jira Service Management, Freshdesk). The workflow engine, SLA tracking, and assignment routing stay. The analytics add-on was replaced.

Monitoring stack (Datadog, New Relic). Real-time infrastructure monitoring with alerting stays. The status page (communication layer) was replaced.

Endpoint security (CrowdStrike, SentinelOne). Hardened, constantly updated security tools with threat intelligence. You do not build your own endpoint protection.

Cloud infrastructure (AWS, Azure, GCP). Production hosting and compute. Railway hosts small internal tools, not production workloads.

Backup and disaster recovery. Data integrity guarantees and regulatory compliance. This stays.

Accounting software (for the freelancer). QuickBooks, Xero, whatever manages your books and taxes. Financial record-keeping is a compliance requirement, not a buildable project.

The pattern is clear. What stays are systems of record, compliance tools, deeply networked platforms, and infrastructure with years of irreplaceable data and integrations. What goes are point solutions, reporting layers, template platforms, and single-purpose add-ons.

The first category earns its subscription. The second category was charging rent for problems that no longer require landlords.

The Time Ledger

Dollars are one measurement. Time is the other.

Each department chapter estimated the hours returned by eliminating manual work and tool management overhead. Here are those numbers consolidated.

That is sixty 40-hour weeks. More than a full-time employee's annual working hours. Recovered not by hiring someone new but by eliminating tool management, manual formatting, data entry, repetitive generation, and the back-and-forth of answering questions that should have been self-service.

For the freelancer, 200 hours per year at an effective rate of $110/hour represents $22,000 in potential billable work. The time savings often exceed the dollar savings. Alex replaced $7,320 in subscriptions but recovered $22,000 worth of billable capacity.

The Honest Limits

The numbers above are real, but they come with conditions. Here is what the tally does not capture.

The Real Shift

Here is what the numbers do not fully convey.

For the past decade, the default answer to every workplace problem was "find a SaaS tool that solves it." Need to track something? Subscribe to a tracker. Need to generate something? Subscribe to a generator. Need to visualize something? Subscribe to a dashboard builder. Each subscription seemed small, $50 a month here, $200 a month there, but they accumulated into a stack that nobody audited and everybody accepted.

The people in this book, Maya, Priya, James, Raj, Lisa, Alex, did not set out to eliminate SaaS. They set out to solve specific problems and discovered that the tools available to them in 2025 could build solutions that used to require vendor subscriptions. They did not become developers. They became people who could describe what they needed to an AI tool and get back something that worked.

That is the shift. Not from one set of tools to another. From being software consumers to being tool builders. From accepting whatever features a vendor decided to ship, to specifying exactly what you need and watching it get built in minutes.

The marketing manager who builds her own dashboard does not just save $3,600 a year on Databox. She gains the ability to modify that dashboard whenever her needs change, without filing a feature request, without waiting for a product roadmap, without upgrading to a more expensive tier. The dashboard does what she wants because she described what she wanted.

The HR manager who builds his own performance review template engine does not just save $4,800 a year on Lattice. He gains review templates that are actually personalized to each employee, something the platform never delivered despite its premium pricing. The tool is better because it was built for his team, not for every team.

The freelance consultant who builds her own monitoring stack does not just save $2,400 a year on Uptime Robot. She gains SLA compliance reports that no monitoring tool offered at her price point. The competitive advantage is the customization, the thing that SaaS products cannot provide because they are built for the median customer.

This pattern repeats across all sixty use cases. Save money. Gain fit. Lose the feature bloat you were never using. Gain the specific capability you actually needed.

That is not an optimization. It is a reclassification. These people stopped being users and became owners.

The Net Number

For a mid-size company running all five departments:

For the freelance consultant:

One Use Case This Week

You are not going to do all sixty use cases. Nobody is. The company that implements every replacement in this book across every department in a single quarter does not exist, and if it did, the change management alone would be its own chapter.

Here is what actually works.

Pick one department. Yours. Pick one use case. The one that made you think "I am paying for that right now, and I barely use it." Pick the one where the SaaS subscription renewal date is coming up. Pick the one where you just had a frustrating experience with the tool last week. Pick the one where the savings number made you pause.

Build it. Not tomorrow. Not next quarter when the budget cycle resets. This week.

The first one takes the longest. You are learning the tool, learning the prompt structure, learning how to describe what you need in a way that produces useful output. The first build might take two hours instead of thirty minutes. That is normal.

The second build takes half the time. By the third, you are not thinking about the tool at all. You are thinking about the problem, describing the solution, and watching it materialize.

Each tool you build replaces a line item on your SaaS budget. Each cancellation creates space, in the budget, in the workflow, and in the mental overhead of managing yet another login, another dashboard, another vendor relationship.

The $47,000 average we started with in Chapter 0? That number sits in your stack right now. Most of it unaudited. Most of it auto-renewing. Most of it solving problems that six AI tools costing $270 a month can solve better, because they solve them for you specifically, not for every company on earth.

The question is not whether these tools can replace your SaaS stack. You have read six chapters of specific, named, priced replacements. The evidence is in the tables.

The question is whether you will open your billing page and start counting. Because every month you wait is another month of paying for tools you barely use, solving problems that are not actually that complicated, at prices that no longer make sense.

Your move.