The short answer
Klaviyo's real weaknesses are mostly mechanics, not missing features: active-profile billing that charges for people you never email, default attribution settings that read generously until configured, frequency guards that drop messages silently, an uneven AI layer, and US-dollar billing that surprises UK brands. Everboost runs Klaviyo daily across client accounts, still recommends it for 7-8 figure DTC brands, and manages every one of these weaknesses deliberately. The platform is excellent; the defaults are not your friend.
First, the disclosure that changes how to read this
Everboost builds and runs Klaviyo programmes daily across its client base, works within Klaviyo’s partner ecosystem, and still picks Klaviyo as the default platform for 7-8 figure DTC health and wellness brands. That is the vantage point here: not a competitor hunting switchers, not an aggregator monetising comparisons, but an operator whose job depends on knowing exactly where this platform bites.
That matters because of who writes the other answers to this question. Nearly every page ranking for Klaviyo review queries is published by a company selling an alternative, and the most-cited source for “Klaviyo’s weaknesses” is a competitor’s comparison page. Those pages are not wrong about everything; they are structurally incapable of ending anywhere except “consider switching”. This one ends somewhere else: run it well.
Weakness one: the billing mechanics, not the price
The headline price gets the attention; the mechanics move the invoice. Klaviyo bills on active profiles, meaning anyone who could be emailed, including people who typed an email at checkout and never opted in. Cleaning your list does not cut the bill by itself: the account must be manually downgraded after suppression, an exception most brands discover after months of overpaying. And the tier bands are uneven enough that sitting just above 10,000 profiles is the most expensive place on the ladder.
None of this is hidden; all of it is documented and manageable. The full mechanics, the band cliffs and the honest cleaning sequence are in our Klaviyo pricing guide. The mitigation takes an hour a quarter. The weakness is that the defaults quietly favour the vendor until someone does it.
Weakness two: attribution reads generously until configured
Out of the box, Klaviyo’s attributed revenue counts Apple’s automated opens as engagement, lets a merely delivered SMS claim an order within 12 hours, includes the full value of orders later refunded, and credits subscription renewals to routine sends. A brand reporting those defaults to a finance team is heading for an argument it will lose.
The fair counterpoint: Klaviyo ships more correction controls than almost any competitor, including bot and Apple-open exclusions and an in-product model comparison tool, and most accounts never open them. Our full breakdown of how the attribution window works and what settings to run covers the honest configuration. The weakness is the default posture, not the capability.
Weakness three: the frequency guard drops messages silently
Smart Sending, the feature that stops customers being over-messaged, has one behaviour almost nobody expects: a blocked message is skipped permanently, never rescheduled, and a skipped flow profile continues without it. A welcome offer can lose a collision with a campaign send and simply never arrive, on exactly the day your calendar was busiest. A/B tests starved by it are cancelled, not delayed.
The mitigation is deliberate configuration per message and a monthly look at the skipped tabs, covered in our guide to when Smart Sending should be on and off. Klaviyo’s own documentation is candid about the mechanics; the weakness is that nothing surfaces the cost unless you go looking.
Weakness four: the AI layer audits better than it creates
Klaviyo’s 2026 AI push is real but uneven. Composer, its AI marketing agent, is genuinely useful for technical audits and quick plain-text campaign volume, and its flow audits catch collisions humans miss. But Klaviyo’s own documentation concedes content generation is weaker on image-heavy templates, which in practice means brands with a developed visual identity will not ship its output. The MCP server makes reporting conversational and audits fast, but flows are read-only on every AI surface, segments cannot be created, and one connection binds to one account, which complicates agency and multi-store setups.
Used for diagnosis, reporting speed and volume, the AI layer earns its place. Expecting it to carry creative or strategy is where disappointment lives.
Weakness five: UK brands pay in dollars, plus the quiet extras
There is no GBP billing. UK brands pay in US dollars through whatever FX spread their card adds, VAT lands on top for non-registered businesses, and UK SMS costs roughly four and a half times US toll-free rates on Klaviyo’s own calculator, so any SMS business case built on American benchmarks quietly fails in Britain. Support on the free plan also expires after 60 days. Details and the sterling maths are in the pricing guide.
What the criticism list leaves out
An honest weaknesses post owes the other column. Klaviyo’s data model and segmentation are the strongest in the category: profile-level behavioural history queryable in ways competitors cannot express. The ecommerce integration ecosystem is the deepest available. The correction controls that featured above as mitigations are themselves a strength most alternatives lack entirely, and the reporting recompute that makes historians nervous also means fixes apply retroactively. These are the reasons the platform remains our default recommendation despite everything above, and why, in our experience, most “should we leave Klaviyo” conversations turn out to be settings conversations wearing a migration costume.
So is it worth it?
For the brands Everboost works with, 7-8 figure DTC businesses whose growth depends on repeat revenue, yes, on one condition: someone operates it. Klaviyo on defaults bills generously, reports generously and skips silently. Klaviyo operated deliberately, with the billing mechanics managed, attribution configured honestly, frequency guards audited and the AI pointed at diagnosis, is the best platform in its category and the one we would choose again.
That condition is the real finding. The platform’s weaknesses are not reasons to avoid it; they are the job description of whoever runs it.
Frequently asked questions
What are Klaviyo's main weaknesses?
Five stand out in daily agency use: billing mechanics that charge for every emailable profile and require a manual downgrade after list cleaning; default attribution that counts Apple's automated opens, delivered SMS and refunded orders until you tighten it; Smart Sending frequency guards that skip messages permanently and silently; an AI layer that audits well but designs poorly and cannot build flows; and US-dollar-only billing with VAT added for non-registered UK businesses.
Is Klaviyo worth it?
For 7-8 figure DTC brands where repeat revenue matters, usually yes: the data model, segmentation depth and ecosystem are genuinely the strongest in class, which is why Everboost deploys it as the default for client programmes. The honest test is whether returning-customer revenue grows faster than platform cost. For very small lists or simple newsletter needs, cheaper tools do the job.
Why do most Klaviyo reviews contradict each other?
Check who wrote them. Almost every page ranking for Klaviyo review queries is published by a company selling a Klaviyo alternative, a review aggregator monetising comparisons, or a Klaviyo agency with the opposite incentive. The most-cited source for Klaviyo's weaknesses is literally a competitor's comparison page. Read any platform review for the facts it cites, never the verdict it reaches, and weight practitioners who disclose their position.
Is Klaviyo expensive?
It is priced above most alternatives at the same list size, but the real cost story is mechanics: you pay for active profiles (anyone emailable, including people who never opted in at checkout), tier cliffs mean crossing 10,000 profiles raises the price sharply, and cleaning your list does not reduce the bill until you manually downgrade. Managed deliberately, cost per revenue is usually strong; left on defaults, the invoice drifts upward.
Are Klaviyo's revenue numbers reliable?
They are a claims model, not a ledger, and the defaults are generous: attributed revenue counts Apple's automated opens, orders after a merely delivered SMS, full order value including later-refunded orders, and subscription renewals. Klaviyo ships more correction controls than most platforms; configured honestly, the number becomes one finance will accept. Shopify remains the ledger.
What are Klaviyo's hidden costs for UK brands?
Billing is in US dollars with no GBP option, so your real cost includes card FX fees; VAT applies (reverse charge if registered, added if not); and UK SMS costs roughly four and a half times US toll-free rates on Klaviyo's own calculator, so US-modelled SMS economics do not transfer.
Is Klaviyo's AI any good?
Uneven in useful ways. The audit surfaces are genuinely strong: Composer's flow audits catch structural errors humans miss, and the MCP server makes reporting conversational. The creative surfaces are weaker: Klaviyo's own documentation concedes content generation struggles on image-heavy templates, and neither Composer nor the MCP can build or edit flows. Use the AI for diagnosis and volume, not for brand-defining creative or strategy.
What does Klaviyo's Smart Sending actually do wrong?
Nothing wrong exactly, but something widely misunderstood: when it blocks a message for frequency, that message is skipped permanently, never rescheduled, and a skipped flow profile just moves on. A welcome offer skipped because a campaign went out that morning is revenue that quietly never happens, and A/B tests that get no results because of it are cancelled outright. The skipped tabs in your reports are worth checking monthly.
Who should not use Klaviyo?
Brands under roughly 5,000 profiles watching every pound, where per-profile cost bites hardest and simpler tools cover the need; content businesses sending simple newsletters; and teams unwilling to manage settings, because Klaviyo on defaults reports generously, bills generously and skips silently. It rewards operators.
What is Klaviyo genuinely best at?
The data model and segmentation: profile-level behavioural data from your store, queryable in segments that other platforms cannot express, plus the deepest ecommerce integration ecosystem in the category and correction controls (attribution settings, bot and Apple-open exclusions, a model comparison tool) that most competitors do not offer at all. The platform's ceiling is very high; its defaults are set for its benefit, not yours.
Does Klaviyo lock you in?
Less by contract than by accumulation: historical analytics, segment logic and flow architecture take work to rebuild elsewhere, and changing attribution settings recomputes your entire reporting history with no frozen record, so your numbers live inside its model. Export cohort reports quarterly and document your settings; portability is a habit, not a feature.
Should you switch away from Klaviyo to save money?
Run the mechanics first: suppress genuinely disengaged profiles after a sunset flow, manually downgrade the tier, tighten attribution so you know what the platform actually drives, and re-check cost against attributed contribution. In Everboost's experience most switch conversations are really settings conversations; brands that migrate to save a few hundred pounds a month usually rebuild the same programme on software with fewer controls.