Ask buyers who run a product analytics tool what frustrates them and the same six complaints come back, whatever the tool. The most common is about the data itself: events that have to be tagged by hand, or autocapture that buries the signal, and numbers that don't quite agree. Below it sit tools only a few people can drive, analysis that runs out before a BI tool would, pricing that climbs as usage grows, data stranded from the rest of the stack, and outages with slow support. What differs is which tool earns which complaint most, and whether a complaint is lived with or acted on. Mostly, buyers live with them — the exceptions are cost and broken data. For the ratings and the pre-purchase view, see what product-analytics and session-replay buyers wish they'd known.
The six complaints — graded
Each complaint carries a verdict: something you live with, something an owner and some discipline can fix, or the kind that ends a contract. The grade is the useful part — every tool here draws complaints, and the question is which of them are dealbreakers for your team.
Tagging is hard, and the data is hard to trust
The most common complaint, and it lands on every tool. On Amplitude and Mixpanel, the burden is manual: “every time we want to add an event, that has to be a code change and our tech team needs to implement that,” says one Amplitude buyer, and Mixpanel buyers say the same about tagging every event in code. Heap's autocapture removes that step and draws the mirror-image complaint — one analytics director says “the auto capture leads people down the wrong path a lot of the time.” Either way, the result buyers describe is numbers they don't quite believe: “depending on where you look, it's a different number,” says a Mixpanel buyer. It is fixable, and buyers often say whose fix it is — “Mixpanel is only as good as our event taxonomy,” says one, and others blame their own setup before the tool. It turns fatal when the doubt never resolves: one company is replacing Heap with Amplitude over blind spots in its data, and another left Amplitude for Heap to get point-and-click tracking. Our recommendation, not a practice buyers describe: name an owner for the event taxonomy before you sign, whichever tool you pick.
Named for this: Amplitude · Mixpanel · Heap · PostHog · Pendo
Only a few people can get answers out of it
The complaint about who actually uses the tool. Buyers describe powerful software that most of the team can't drive: “I think that the tool is really powerful, but I really don't know how to use it,” says one Heap buyer, and a Mixpanel buyer says business teams find it “difficult for them to adopt because they can't quite figure out how to use it.” On Amplitude it shows up as training — “the learning curve can be quite steep, and it does take training to get them up to speed” — and on PostHog as a tool built for people who know SQL and the warehouse. The result is a tool bought for many and used by a few: “Basically, not all the users are using it,” says an Amplitude buyer. It is fixable, but the fix is organizational: buyers who describe low adoption usually describe no one owning the tool's rollout. The exception is Pendo, whose users in these interviews do not raise it.
Named for this: Amplitude · Mixpanel · Heap · PostHog
It stops short of what a BI tool or the warehouse can do
The complaint buyers mostly work around. Product analytics tools are praised for funnels, retention and behavior, and criticized when asked to be the whole reporting stack: one buyer calls Amplitude “a good tool for product analytics, but it's really not strong for e-commerce,” and another says its reporting “is starting to fall behind” what the team gets in its warehouse. Pendo buyers say there are things the team can't track there, including business metrics, and Heap buyers name metrics it can't calculate. Buyers raise this limitation across the five tools, but rarely describe it as the sole reason for switching. Instead, they often keep product analytics for funnels, retention and behavioral analysis while relying on a BI platform or warehouse for broader business reporting.
Named for this: Amplitude · Mixpanel · Heap · Pendo · PostHog
Pricing climbs as usage and requirements grow
The complaint that most often ends a contract. Buyers describe several reasons analytics bills grow: data volume, expanding usage and additional modules. An Amplitude buyer describes volume-based pricing — it “is priced, if I remember correctly, based on volume” — while a PostHog buyer says the tool starts cost-effectively before costs rise “as your data needs expand.” Other buyers describe paying more for additional capabilities. Unlike the complaints above, this one produces full switches: one company replaced Amplitude with a cheaper tool and accepted rougher reporting to pay less, another moved from Pendo to PostHog — “The main driver was money” — and a third passed on Amplitude's testing module for a cheaper option. For the cost mechanics themselves, see what buyers actually pay for analytics.
Named for this: Amplitude · Pendo · PostHog · Heap · Mixpanel
The data won't join up with the rest of the stack
The complaint about everything around the tool. Buyers describe product data stranded from revenue and customer data — “So data lives in Shopify, and it lives in Looker, and it lives in Amplitude,” as one product leader puts it, and a PostHog buyer names “not being able to connect our Stripe information” as the weakness that remains. Others describe identity that breaks before a purchase and exports that are hard to get out. It is fixable, and the fix buyers describe runs through the warehouse: one company moved its audience activation from Amplitude to a warehouse-native tool while keeping Amplitude for product analytics.
Named for this: PostHog · Pendo · Amplitude · Heap · Mixpanel
Outages and slow support when the data breaks
The least common complaint, and a serious one when it lingers. When a connector fails or a script loads late, buyers lose data they can't recover, and slow support turns that into lost trust. One product leader says Heap has had outages “that have been pretty crippling for us,” and an Amplitude buyer on Shopify says the script “loads really late, and we lose a good chunk of traffic.” Support splits buyers rather than uniting them — Heap's draws both praise and complaint in these interviews. But when recurring data problems or outages cannot be resolved promptly, the consequences can extend beyond inconvenience: one company replacing Heap with Amplitude describes both data blind spots and slow resolution.
Named for this: Heap · Amplitude · PostHog · Mixpanel
Want this read against your own stack?
Get my read →The stories behind the complaints
The counter-current: what buyers don't complain about
Two complaints you might expect are mostly missing. It tells you what happened but not why almost never comes up as a grievance — buyers who want the why simply run a session-replay tool alongside, and one says they go to Hotjar to validate a hypothesis they first see in Amplitude. And AI features split buyers rather than frustrate them: a few Amplitude users call its AI lagging or inaccurate, while others praise it, and Heap's users mention its AI only to praise it. The tagging complaint is also less about any vendor than about a method — buyers move from hand-tagging to autocapture and back again, and the complaint follows them. Before you conclude the tool is the problem, separate what the vendor controls — such as pricing and support — from what depends on the tracking approach and your own implementation. The latter may be fixable without switching, but a different platform can still make sense when its approach better fits your team's resources and requirements.
What this means — whether you run one or you're buying one
If you already run a product analytics tool, sort your complaints with the grades above. The livable one, limited depth, is answered by putting a BI layer or the warehouse beside the tool. The fixable ones — tagging and data trust, low adoption, data that won't join up — are about ownership: an event taxonomy someone maintains, a rollout someone owns, and a warehouse the tool feeds. The ones buyers act on are pricing that outgrows the value at renewal and data that stays broken when support is slow. If you're evaluating one, the trade to decide first is tagging: engineering effort up front with hand-tagged events, or cleanup and trust work later with autocapture. Then price the tool at the data volume you expect in two years, not the one you have today. Both are our recommendations rather than practices buyers describe. For the head-to-heads, see Amplitude vs. Mixpanel, Amplitude vs. Heap and Mixpanel vs. Heap.
Common questions
What are the most common complaints about product analytics tools?
Six recur across Amplitude, Mixpanel, Heap, PostHog and Pendo. The most common is data that's hard to trust because of event tagging — events tagged by hand in code, or autocaptured noise. Next come tools only a few people can use, analysis that stops short of what a BI tool or the warehouse can do, pricing that climbs as usage grows, data that won't join up with the rest of the stack, and outages or slow support when the data breaks. Most buyers live with them; cost is the complaint most likely to end a contract.
Is autocapture better than manual event tagging?
Neither escapes the complaint. Buyers on Amplitude and Mixpanel say every new event is a code change their engineers have to make, and one company left Amplitude for Heap to get point-and-click tracking. Heap's autocapture draws the opposite complaint, data that leads people the wrong way, and one company is replacing Heap with Amplitude over blind spots in its data. The trade is engineering time up front against cleanup and trust later.
Why do companies leave a product analytics tool?
In these interviews, mostly over cost. The clearest full switches were Amplitude to a cheaper tool and Pendo to PostHog, both on price, and one company passed on Amplitude's testing module for a cheaper one. Data that couldn't be trusted, plus slow support, drove one company from Heap to Amplitude. Most other complaints — tagging, depth, the learning curve — are lived with rather than acted on.
Are these complaints reasons to switch, or can you live with them?
Mostly the second. The tagging and data-quality complaint is usually part of the buyer's own setup, and buyers say so; missing depth is handled by pairing the tool with BI or the warehouse; and a tool few people use is an ownership problem before it is a product one. The complaints that end contracts are pricing that outgrows the value at renewal, and data that stops being trusted when outages or slow support leave it broken.
Get this research made for your stack
Running Amplitude, Mixpanel, Heap or another product analytics tool and wondering whether your complaints are normal? Do a 15-minute interview about your tracking, your team and your stack, and get this personalized: which of these complaints peers on your tool share, and which ones end with a replacement.
Get my personalized read — 15-min interviewNo password · your interview is anonymized before it ever informs a page like this one.