The Subscription Tax: Why B2B Software is Breaking for Small Teams

The default pricing model in B2B software is simple: pick a tier, pay every month, forever. It works fine if you're Salesforce selling to a Fortune 500. It works very poorly if you're a five-person company that hires someone once every six months and is being asked to pay $300/month for the privilege of having an ATS sit there waiting.

Call it the subscription tax: the gap between the value you actually extract from a tool and the price you pay to keep it on retainer for the eleven months a year you're not using it. For small teams, that gap has gotten absurd. This is a manifesto for fixing it.

Run the math, it's worse than you think

Let's pick a realistic small-team profile. You're a 5–15 person company. You hire roughly twice a year. Each hire involves screening 100–300 resumes, running maybe 12 interviews, and making one offer. Total active “ATS-using time” per role: about three weeks.

Now compare two pricing models for that profile.

Model A: A typical mid-market ATS subscription

• $249/month base, “Growth” tier (the one that actually has the AI features)
• $49/month per active user × 3 users = $147/month
• Annual contract, no monthly option
• True annual cost: $4,752
• Cost per hire (2 hires/year): $2,376
• Cost per active week of usage (6 weeks): $792/week

Model B: Transactional, pay-per-job pricing

• ~$50–$100 per job posting, includes the screening, ranking, and storage for that role
• No seats, no base fee, nothing recurring
• True annual cost (2 hires): $100–$200
• Cost per hire: $50–$100
• Cost during the 46 weeks/year you're not hiring: $0

That's not a markup. That's a different category of cost. And the worst part is you don't feel it as one big charge — you feel it as a $396 line on a credit card every month for years, easy to ignore, easy to forget you're paying.

Why subscriptions became the default (and why it's breaking)

Subscription pricing took over B2B in the 2010s for good reasons. It smoothed revenue for vendors, made budgeting predictable for buyers, and aligned with how a typical 200-person company actually used software (continuously, with everyone logged in every day). The model fit the customer.

That fit has eroded. Today the buyer pool for “an ATS” includes:

• Ten-person SaaS startups that hire 3–4 times a year
• Local agencies hiring once or twice a year
• Solo founders making their first hire ever
• Side-project teams who need a real hiring tool for a single role and then nothing for two years

For all four of those, paying every month is a tax, not a service. And the rise of zombie subscriptions — recurring charges nobody on the team remembers signing up for or feels empowered to cancel — is starting to genuinely show up in small-business P&Ls. A recent wave of finance-ops tools (Vendr, Spendflo, even the “cancel subscriptions for me” category) exists almost entirely because of this problem.

The bloat tax: what you're paying for that you'll never use

Subscription pricing also has a perverse incentive on the vendor side: the only way to justify the next price tier is to keep adding features. So enterprise ATS platforms now bundle in features designed for HR departments at companies ten times your size, and you get to pay for them. Some of the standard inclusions:

• Career page builders — you have a website. You don't need a worse, slower one hosted at a different domain.
• DEI reporting dashboards — useful at 500 people, theatre at 5.
• Complex permission systems — hiring managers, recruiters, coordinators, sourcers, approvers… your team is four people who all trust each other.
• Approval workflows for offer letters — the “workflow” is the founder typing the offer in Google Docs.
• HRIS integrations — you don't have an HRIS.
• Job-board syndication to 80 sites — you're going to post to LinkedIn and one or two niche boards. The other 78 are spam farms.
• Onboarding modules — that's a different product the same vendor wants to upsell you on next quarter.

None of that is bad software. It's just not your software. You're renting a 12-bedroom mansion to sleep in the guest room.

What a small team actually needs (the checklist)

Before you sign the next ATS contract, screenshot this list and check the proposal against it. If a feature's on the “need” column, you should be able to point to exactly where in the product it lives. If it's on the “don't need” column, you shouldn't be paying for it.

Need (the actual job)

1. Resume intake — accept resumes by upload, email, or a public job link. No more.
2. Ranking that explains itself — sort candidates by fit and tell you why in plain language, so a non-technical screener can act on it.
3. A shared shortlist — the founder and the hiring manager need to see the same 10 names with the same notes.
4. Candidate communication — at minimum: send rejection emails in a batch, schedule a screening call, request a follow-up document.
5. Data security & deletability — encrypted at rest, exportable, and deletable on request. This is non-negotiable.
6. An exit door — you can export everything (resumes, notes, scores) on the day you stop using it. No data hostage-taking.

Don't need (the bloat tax)

• DEI reporting dashboards (until you're 50+ people)
• Five-tier permission systems
• Career page builder hosted on a vendor subdomain
• Job-board syndication to 80 sites you've never heard of
• Offer-letter approval workflows
• Onboarding / e-signature / payroll bundling
• An “analytics suite” for a pipeline of 12 people
• Required annual contracts

The un-subscription manifesto, in three lines

1. Pay for the thing when you use the thing. Recurring charges should match recurring usage.
2. Bundle is a tax. If you're paying for nine features to use two, you're subsidizing somebody else's product strategy.
3. Zero is a valid price for zero usage. The default cost of an idle tool should be $0, not “your annual contract.”

Subscription B2B software isn't evil — it's just been mispriced for small teams for over a decade. The good news: the market is finally catching up. Pay-per-use, per-job, per-project pricing is back in fashion, and tools designed around it (including ours) are starting to win on the only metric that matters for a five-person team: how much you pay per actual hire, not per calendar month.

Audit your stack this quarter. Cancel what you're not using. And the next time a vendor quotes you a per-seat-per-month price for software you'll touch six weeks a year, do the math from this article on the back of the napkin in front of them. Watch the conversation change.