The honest problem with review sites

It seems like most "best ATS" lists are actually "most marketed ATS" lists - ranked by which vendors invest most in review generation campaigns, which companies have the largest sales teams to push customers toward review platforms, and which brands have the most marketing budget for SEO dominance, not which products actually deliver the best outcomes for companies in your specific situation.

That's a strong claim. Here's why it's accurate:

G2 and Capterra rankings reward volume of reviews. Volume of reviews correlates directly with: how large the vendor's customer base is, how aggressively the vendor's customer success team requests reviews, and how much budget the vendor invests in review generation campaigns. None of these correlate reliably with product quality for your specific use case. The ATS with 4,200 reviews from a 10,000-customer enterprise platform is not necessarily better than the ATS with 180 reviews from a 500-customer focused platform - it just has more reviews, for reasons unrelated to product quality.

Enterprise ATS vendors dominate rankings despite serving a completely different use case. iCIMS, Greenhouse, and Workday have thousands of customers and thousands of reviews. A significant proportion of those customers are mid-market or enterprise companies. A 50-person company evaluating ATS options reads "iCIMS: #1 ATS by market share" and interprets that as a recommendation - when in reality, iCIMS was built for and prices for companies 10-100x their size.

SEO-driven "best ATS" articles are content marketing, not buyer research. The articles ranking highly for "best ATS 2026" on Google are almost uniformly produced by companies with affiliate relationships with the ATS vendors they recommend, or by the ATS vendors themselves through their content marketing. They are not written by people who have evaluated the tools - they are written by content teams optimising for keyword rankings. Reading them tells you which vendors invest in content SEO. It tells you very little about which ATS is right for your company.

What actually matters: the selection criteria review sites can't see

The meaningful differentiators in ATS selection are almost entirely invisible to review site rankings. They're only visible through a specific evaluation process - and specifically through conversations with customers who match your profile.

Hiring volume is the primary segmentation criterion

An ATS built for a company hiring 500 people per year in retail and logistics (Paradox, Jobvite, iCIMS) is a completely different product than an ATS built for a company hiring 30-80 people per year in professional services (Greenhouse, Lever, Ashby). Both are "enterprise" on some review sites. Both have positive reviews. Neither is the right choice for the other's context.

The hiring volume ranges that map to genuinely different product categories:

  • Under 20 hires per year: Simple, affordable, fast-to-implement ATS. Primary concern: ease of use, fast time-to-value. Ashby is too complex; iCIMS is architecturally and financially inappropriate.
  • 20-100 hires per year: The most contested market segment. Mid-market ATS platforms (Greenhouse, Lever, Workable, Teamtailor, Treegarden) compete here. Feature depth, integration capability, and pricing transparency differentiate.
  • 100-500 hires per year: The premium mid-market and lower-enterprise range. More complex pipeline management, more integration requirements, more reporting depth needed. Greenhouse and Ashby compete well; iCIMS is entering the frame.
  • 500+ hires per year in standardised roles: Enterprise territory, often with high-volume flavour. iCIMS, Workday, SAP SuccessFactors, Paradox (agentic AI).

The 3 must-have filter

Is it crazy to start an ATS evaluation with a 40-feature checklist? Not crazy - but consistently counterproductive. The checklist approach to ATS selection produces over-buying, implementation complexity, and eventual underutilisation of features the company paid for but doesn't actually need.

The more effective approach: identify the 3 features that are genuinely non-negotiable - without which the platform cannot serve your core hiring workflow. Not "nice to have." Not "we'd use it if it's there." The 3 features whose absence would make the platform non-functional for your primary use case.

For most mid-market companies, those 3 features are something like: reliable CV parsing and structured candidate records, calendar integration for interview scheduling, and job board posting to the specific platforms your candidates use. Everything else is secondary - it may influence your decision between two equally capable platforms, but it shouldn't drive the selection.

Starting from 3 must-haves rather than 40 nice-to-haves produces a shorter shortlist, faster evaluation cycles, and better decisions.

Pricing transparency as a filter, not a data point

If you can't see the price before you talk to a sales rep, move on. That's a stronger filter than it might sound, and here's why:

An ATS vendor with opaque pricing has made a deliberate strategic decision: they believe they can extract more revenue by quoting each customer individually (and adjusting the quote based on the customer's perceived willingness to pay) than by publishing a price that everyone can see. That decision reveals something about the vendor relationship you're entering. It means the price you're quoted is not the price your competitor in a comparable situation is paying. It means the renewal price is undiscoverable. It means the negotiation is structural and ongoing, not transparent and predictable.

The vendors with transparent, public pricing - where you can see what you'll pay before booking a demo - have made the opposite decision. They've chosen to compete on product quality and clear value rather than negotiation skill. That alignment of incentives tends to produce a better long-term vendor relationship.

Treegarden is an example of a platform in the 50-750 employee range with fully public pricing: Startup at $299/mo, Growth at $499/mo, Scale at $899/mo (GBP £235/£395/£710). It's not the only option with transparent pricing, but it's one example of what that looks like as a design principle.

The real test: your workflow, your data

The single most diagnostic evaluation step that review sites cannot replicate: run a real job and two real CVs through the system during the vendor demo, not in a guided tour of the vendor's prepared demonstration.

Provide the vendor with a copy of your most common job description and two actual CV files (appropriately anonymised) before the demo. Ask them to start the demo by parsing those CVs and walking through how a recruiter would process those candidates in their system. The output will tell you:

  • How accurately the AI parses your actual CV format
  • What the recruiter experience actually looks like for a workflow you recognise
  • How the system handles the edge cases in your real data (unusual formatting, non-standard skills, etc.)
  • Whether the platform's assumptions about how hiring is done match your actual process

A vendor who declines to run your data and insists on a scripted demo is showing you something important about their confidence in product performance on real-world inputs.

The reference check that review sites replace - badly

Review sites exist, at least partially, because the alternative - talking to actual customers - takes time. But the time investment in 2-3 focused reference calls is consistently the most valuable evaluation step. Here's how to get the most from it.

Ask the vendor for two references with these specific characteristics: in your industry (or similar), your company size range (within 50% of your headcount), and who switched from your current ATS (if applicable) or who started from a comparable alternative. These specifics matter because a reference from a 2,000-person enterprise company who loves the platform is not informative for a 100-person company's evaluation.

In the reference call, the most valuable questions are not "do you like the platform?" (the answer is always yes - the vendor has pre-selected the reference) but rather: "What did you underestimate about the implementation?" "What do you wish you'd known before signing?" "How has the vendor handled problems - give me a specific example of something that went wrong and how they responded." "How did the renewal conversation go - did pricing change?"

The last question is the one most buyers forget to ask, and it's the one that reveals the most about the vendor relationship you're entering.

Ask about renewal pricing before signing - always

The best time to understand your renewal pricing risk is before you sign the initial contract, not 11 months later when the renewal notice arrives.

Ask specifically: "What is your standard renewal pricing policy? Is there a cap on annual increases in the contract? Can you show me what a typical customer at our size paid at their last renewal compared to their initial contract price?"

Vendors who have a good renewal story will answer this readily. Vendors with aggressive renewal histories will redirect, qualify, and deflect. The quality of the answer is informative regardless of the specific number.

See exactly what Treegarden costs

All features included. Public pricing. No demo required to see the numbers. Startup: $299/mo equivalent · Growth: $499/mo equivalent · Scale: $899/mo equivalent.

View full pricing →

Frequently asked questions

How do I find the best ATS for my company?

Reject popularity as a selection criterion. Define your hiring volume, identify your 3 genuinely non-negotiable features, filter to platforms with transparent pricing, test with your own data during the demo, request references specifically from companies in your industry and size range, and ask about renewal pricing policy before signing. This process consistently produces better outcomes than starting with review site rankings.

Are G2 and Capterra reviews reliable for ATS?

Useful for identifying consistent complaints across many reviews, understanding the general customer relationship tone, and getting a sense of onboarding experience. Unreliable for comparing vendors across different company size segments, comparing vendors with very different review volumes, or making "best" determinations. Any "leader" badge is a volume-based or paid placement, not a quality assessment for your specific context.

What's the difference between a popular ATS and the right ATS?

A popular ATS has the most users and the most marketing. The right ATS fits your specific hiring volume, team structure, industry, must-have integrations, and 3-year budget. iCIMS holds the largest share of the ATS market according to APPS RUN THE WORLD (2025) - it is right for some companies and completely wrong for others. Popularity tells you the vendor has successfully sold the product; fit tells you whether it will work for you specifically.

How do I evaluate ATS vendors without getting sold to?

Three practices: Control the demo agenda - send a specific list of workflows you want to see, not "show me the ATS." Test with your own data - provide a real job description and two anonymised CVs and ask them to process these during the demo. Ask the question they don't want to answer: "What do customers who cancel typically say as their reason for leaving?" A vendor with good retention answers this specifically. One without that confidence deflects.

Common mistakes companies make when selecting an ATS

Most ATS mis-selections share a recognisable pattern. Understanding the mistakes before you start the evaluation process is more useful than diagnosing them after you've signed a two-year contract.

Mistake 1: starting with the feature checklist instead of the workflow

The instinct to build a comprehensive feature checklist is understandable - it feels rigorous. In practice, it produces a list of 40-80 features, most of which the company will not use or will use rarely. The checklist drives the evaluation toward the platforms with the longest feature lists, which tend to be the largest, most complex, and most expensive platforms. A 60-person company ends up evaluating Workday or iCIMS because those platforms checked more boxes on a list that was never validated against actual hiring frequency.

The correction: start with your three actual hiring workflows - the workflows your team runs every week. Map each workflow step by step. Then identify the 3 features without which those workflows break entirely. Evaluate platforms against those 3 features first, before considering anything else.

Mistake 2: evaluating the demo, not the product

Every ATS vendor has a polished demo environment with clean data, pre-configured pipelines, and scripted flows that highlight the strongest parts of the product. Evaluating the demo is not evaluating the product. It is evaluating the vendor's ability to construct a compelling demo. These are different skills that do not correlate reliably with product quality.

The correction is to control what gets demonstrated. Send the vendor two anonymised CVs from your actual candidate pool and a real job description before the call. Ask them to start the demo from the beginning of your actual intake workflow - not from a pre-loaded pipeline. The quality of parsing, the speed of the workflow, and the clarity of the interface on real data tell you more than any scripted walkthrough.

Mistake 3: letting procurement timelines drive the decision

Procurement pressure creates compressed evaluation timelines. When a hiring manager needs the new ATS running in six weeks because a growth round just closed, the evaluation shortens from 8 weeks to 2 weeks. The evaluation shortcuts in a predictable order: reference calls get cut first (too slow), vendor comparisons collapse to one finalist instead of two (faster), contract review gets minimised (urgency). The result is a worse decision made faster, often locked in for two or three years.

The correction: build a standard ATS evaluation playbook that can run in 4-6 weeks at normal pace, not 2 weeks under pressure. The playbook should have a defined shortlist (3-4 platforms maximum), a standard demo script, a reference call template, and a contract review checklist. Running the playbook at normal pace consistently produces better outcomes than compressing it under urgency.

Mistake 4: ignoring total cost of ownership

The platform subscription price is rarely the total cost of the ATS. The hidden cost elements most commonly missed: implementation and onboarding fees (often $2,000-$8,000 for mid-market platforms and not included in the headline price); integration development (connecting your HRIS, background check provider, or calendar system often requires developer time); data migration from your existing system; training time for hiring managers and HR coordinators; and renewal price increases in years 2 and 3.

The correction: build a 3-year total cost model before comparing platforms. Include the subscription at renewal prices (assume 10-15% increase if the contract doesn't cap it), estimated implementation costs, integration costs, and a rough estimate of internal time investment. The platform with the lowest headline price is rarely the lowest 3-year cost.

For independent guidance on HR technology total cost frameworks, SHRM's guide to selecting HR technology covers cost evaluation components for HR technology purchases, including considerations that apply directly to ATS selection.

A practical ATS evaluation checklist

The following sequence is designed to produce a good ATS decision in 5-6 weeks without requiring a dedicated procurement team. It works for companies with a solo HR manager or a small TA team.

Week 1-2: define requirements and build shortlist

  • Document your 3 core hiring workflows in step-by-step detail (not features - actual workflow steps: "recruiter receives application email, logs into system, opens candidate profile, leaves note for hiring manager, advances candidate to phone screen stage").
  • Identify the 3 features without which those workflows cannot run. Write them down and commit to them as elimination criteria.
  • Set a firm budget range for 3-year total cost including implementation.
  • Build a shortlist of 3-4 platforms that match your hiring volume segment, have publicly available pricing or pricing you can obtain without a sales meeting, and have verified customers in your industry. Do not evaluate platforms outside your volume segment regardless of brand recognition.

Week 2-3: run structured demos

  • Before each demo, send the vendor your two anonymised CVs and one job description. Request that the demo start with parsing those inputs, not from a pre-configured pipeline.
  • Run the same script with each vendor: ask them to demonstrate the same 5-7 workflow steps in the same order. Variation from the script tells you something about the platform's flexibility (or lack of it).
  • Score each vendor on your 3 non-negotiable features immediately after the demo, while the experience is fresh. Score on a 1-5 scale on each feature, not a general impression.
  • Eliminate any platform that scores below 4 on any of your 3 non-negotiable features. Do not advance platforms to the reference stage on the assumption that a weak feature will improve.

Week 3-4: reference calls

  • Request 2 references per remaining platform. Specify: same industry (or similar), same headcount range (within 50%), customer for at least 12 months. Reject references that don't meet these criteria and ask for alternatives.
  • In each reference call, ask specifically: "What did you underestimate about implementation?" "What do you wish you'd negotiated differently in the contract?" "How has renewal pricing worked?" "What's the most frequent complaint your team has about the product day-to-day?"
  • Ask about support response times - "when something breaks, how long does it typically take to get a resolution?" This is a more reliable indicator of support quality than the vendor's stated SLA.

Week 4-5: contract review

  • Read the data portability clause before signing. Can you export all candidate data, pipeline history, and notes in a standard format if you switch platforms? Some ATS vendors make data export difficult by design.
  • Find the renewal pricing clause. Is there a cap on annual price increases? If not, ask to add one - a 5-10% annual cap is a reasonable ask for a 2-year contract.
  • Understand the minimum commitment and exit provisions. What happens if your hiring volume drops significantly in year 2? Is there a pathway to downgrade or pause?
  • Clarify what is and is not included in the stated price: implementation, training, support tier, and integrations. Get exclusions in writing, not in a verbal assurance from a sales rep.

According to LinkedIn's Global Talent Trends research, talent acquisition teams consistently report that process efficiency and technology fit are among the primary factors affecting recruiter productivity - making structured ATS selection a direct investment in recruiting capacity, not just a software purchase.

How ATS needs change as companies grow

A common error in ATS selection is choosing a platform for where the company plans to be in three years rather than where it is today. The logic sounds sensible - buy ahead of growth to avoid re-implementation costs. In practice, it produces platforms that are over-configured for current needs, slow to onboard, and underutilised because the team hasn't reached the hiring volume that justifies the complexity.

The more accurate framework is to choose an ATS that fits your current hiring volume and has a credible upgrade path (either within the same platform or through a well-defined migration process) when you reach the next tier. A company currently hiring 25 people per year should not implement iCIMS or Workday on the assumption that they'll be hiring 500 per year in three years. The implementation cost, complexity, and admin overhead of an enterprise ATS at 25 hires per year are not justified by speculative future volume.

What does change as companies grow, and how ATS needs shift:

  • Under 50 employees: The ATS needs to be fast to configure, simple enough that non-technical hiring managers can use it without training, and priced appropriately for a team where HR may be a part-time function. Reporting is secondary to workflow speed.
  • 50-200 employees: Interview coordination complexity increases. Multiple hiring managers with different preferences, panel interviews, structured interviewing programmes, and role-specific evaluation criteria all emerge at this stage. The ATS needs to accommodate structured feedback, not just pass/fail tracking.
  • 200-500 employees: Analytics and reporting become operationally important, not just useful. Time-to-fill tracking, source quality reporting, and pipeline conversion analysis are needed to manage TA team capacity and forecast hiring timelines. Integration with the HRIS becomes critical rather than optional.
  • 500+ employees: Compliance reporting, approval workflows, and high-volume role management (roles where you're processing 50+ applications in the first 48 hours) require purpose-built features that most mid-market ATS platforms handle poorly or don't handle at all.

The key takeaway: match the ATS to your current stage, not your aspirational stage. The switching cost of moving from a mid-market ATS to an enterprise platform when you actually reach enterprise hiring volume is lower than the operational cost of running an oversized system for three years while you grow into it.

For employer-side data on how hiring technology adoption correlates with company growth stages, SHRM's HR Technology research hub tracks HR technology investment trends and adoption patterns, though it is not ATS-specific.