AiHello review: the bidding is real, the documentation is not

The whole company is one page, and that shapes how you buy it

Most reviews of AiHello read like reviews of an ordinary software vendor. It is not one, and the shape of the website is the first thing a buyer should know, because it determines what evidence is available to you.

Checked again on 21 August 2026. The site is a single page. The navigation items labelled Features and Pricing are anchors that scroll you down that same page. Request a pricing path, a product path, or a URL you invent on the spot, and all three return the identical homepage, because the site runs a catch-all route. The pricing is genuinely there and genuinely public — there is simply no separate page holding it, which is why so many comparison posts cite a URL that does not exist as a document.

Behind that sits a blog, weighted heavily toward competitor-alternative posts and one long roundup of Amazon PPC tools that is doing most of the search work for the entire domain.

None of this makes the software bad, and it is not a criticism of a small team choosing to spend its engineering hours on the product. It does mean the normal evaluation method fails. There is no documentation to read, no feature matrix to check against your edge cases, no changelog to see what shipped last quarter, no security page for your procurement team. Everything you learn, you learn on a call or inside the product. That is a real cost to weigh, and it changes what a sensible evaluation looks like — which is what the rest of this page is about.

What the price is, and why the per-marketplace multiplier decides it

The pricing block on the current page lists, as printed:

• Smart — $199 per month per marketplace, plus 2% of ad spend.
• Advanced features — Repricing and Budgeting, at $10 per SKU per month each. They are separate line items, so price the modules you will switch on rather than assuming one covers both.
• Expert-managed — custom, book a call.

By the standards of this category that is more disclosure than most, and both halves of the fee are stated, which several better-known vendors do not manage. What complicates it is that older AiHello landing pages remain reachable carrying a different ladder — flat monthly tiers banded by ad spend, no per-marketplace multiplier and no percentage. Those legacy pages are where most of the cheaper AiHello figures in comparison posts come from, and a writer who lands on one has no signal that it is superseded. Treat the current block as the live one and get the quote in writing against it.

The multiplier is the part to model rather than skim, and it is easier to see as arithmetic than as a sentence. On a single marketplace, AiHello prices at $199 plus 2%. On four — the United States, Canada, the United Kingdom and Germany, a normal enough footprint — that is $796 of base fee before a single percentage point is counted.

Which produces a concession we are happy to make in public, because a reader who checks the arithmetic and finds it holds will believe the rest of the page. On one marketplace, AiHello is the smaller invoice than Dr. PPC at every spend level up to roughly $115,000 a month, at which point our cap finally crosses their uncapped 2%. On four marketplaces the picture inverts twice: we are smaller below about $49,600 of monthly spend, they are smaller between there and about $85,200, and we are smaller again above that as the cap holds our line flat. The slider below is the honest way to read that, because there is no single answer to give you.

The entire public record is 36 reviews, and the three platforms disagree

When a vendor publishes little about the product, third-party reviews carry more weight than they can bear. So we opened all three primary records on 21 August 2026 rather than quoting an aggregate — and printed the count beside every score, because a score without its count is not a fact.

• G2 — 4.6 across 14 reviews.
• Capterra — 4.2 across 13 reviews.
• Trustpilot — TrustScore 3.3 across 9 reviews.

Thirty-six reviews in total, and a spread of more than a point between the highest and lowest average. We are not going to average those into a verdict, and neither should anyone else. Different platforms attract different populations — one collects invited reviews from software buyers, another collects whatever an unhappy or delighted user goes looking to write — and at counts in the low teens the average describes the collection method more than the product.

The right conclusion is not that AiHello is a 4.6 or a 3.3. It is that no rating anywhere is going to decide this purchase for you, and a page that tells you otherwise has not opened the other two records. Note also that all three sites refuse ordinary requests, which is exactly why most articles quote whichever one they happened to reach and present it as the rating. We have no public review record at all, which is a gap on our side rather than a point in our favour.

What the product does genuinely well

AutoPilot bidding. This is why people buy it. Bids move continuously toward a target and each change is logged, so you can reconstruct what moved and when. The behaviour users single out is the intraday work — pushing bids up in converting hours and down in dead ones. That is real dayparting rather than a scheduler that pauses a campaign at nine o'clock, and it is not universal in this price band.

Bulk campaign creation. Building hundreds of campaigns across match types and targeting types is where the tool visibly earns its fee. If you are launching a wide catalogue this is the strongest single feature on offer, and it is worth the subscription for a launch quarter on its own.

That strength points at a real problem it is well suited to. In one catalogue audit we ran, the products outside a brand's top twenty were receiving about $2,000 of monthly ad spend between them while producing roughly $30,000 of sales, with individual listings funded at $2 to $35 a month on settings nobody had touched in years. That is one account rather than a benchmark, and we are describing it as a mid-market consumer-goods catalogue and nothing more specific. But the shape is common, and it is exactly the kind of neglect a bulk campaign builder plus continuous bidding is built to clear.

Keyword harvesting and negation are automated and competent. A managed option genuinely exists — they employ specialists and sell running the account as a tier, where many software vendors say the same thing and mean a support ticket. And the app is listed in the Amazon Selling Partner Appstore across several regions, so the data connection is a sanctioned one rather than a scraped one.

The limit is one they publish themselves: a bidder is not an analyst

No custom rules. AiHello's own comparison table records this as the trade-off, which is the right place for a criticism to come from. If your account has an idiosyncrasy that needs a specific condition handled a specific way, you are asking a model to infer it. Sellers who want deterministic control tend to be happier with a rules engine.

That is a design choice rather than a defect, and there is a genuine argument on AiHello's side of it. Rules fail silently. On one supplement account we worked in, a single bad line in a rule sheet kept raising bids on campaigns whose ACOS was already too high, and the tool did precisely what it had been told for weeks while reporting as healthy the entire time. Our correction was structural: one senior practitioner built a verified reference sheet, and every other account's rules are now checked against it. A model you cannot edit cannot be misconfigured that way.

But the mirror question survives the switch, and it is the one to put to any automated ad product including ours: who reviews the targets, how often, and against what reference? A model that hits a wrong target with precision is the same failure wearing better clothes. AiHello moves bids toward the goal you set. It does not tell you the goal is wrong for that product this quarter, or that the product should not be advertised at all while its rating sits where it does.

That distinction has teeth, and here is what it looks like at scale. Read out of our own Orbit database across the 47 brands we manage — 3,086,624 search terms that took a click, Amazon search data from 1 May 2026 onward — 48.5% of all search spend went to terms that produced no orders: $4,962,963 of $10,243,379. The efficiency headline hides it in both directions: 6.4% of those brands run under 15% ACoS and 23.4% run above 50%, so a target that reads as healthy on one account is an emergency on the next. A bidder moves toward whichever number you typed in. It does not tell you the structure underneath is the thing that needs changing, and optimising bids on that structure improves the wrong thing, efficiently.

How to evaluate a vendor that publishes no documentation

Since there is no spec sheet to read, the trial has to carry the whole evaluation. Most trials of bidding software are worthless because three things changed at once. Do it deliberately.

• Write down one metric before you start. Contribution after ad spend is the right one. ACOS improves beautifully when you simply spend less, which is not what you are testing.
• Hold back a control group of comparable ASINs and leave them untouched for the duration.
• Change nothing else — no price moves, no new creative, no listing rewrites, no promotions.
• Give it long enough. Our own working rule across managed accounts is that seven- to fourteen-day windows are unreadable, the decision window is four to eight weeks, and an account takes three to four weeks to settle after a spend change. A fortnight measures the disruption of switching, not the tool.
• Subtract the fee, including the percentage and every marketplace base. A tool that improves performance by less than it costs has taken a quarter as well as the money.

Alongside it, ask the four questions the missing documentation would have answered: which ad types the automation actually touches versus only reports on; what happens to bids when a data pull from Amazon fails; whether there is any export of your own historical data if you leave; and what the notice period is. Those are ordinary questions, they are fair to ask of us too, and a small vendor answering them straight on a call is worth more than a large one with a polished page.

Who AiHello suits better than we do

Buy AiHello if you sell on one marketplace with a wide catalogue, want continuous bidding and real dayparting without an enterprise contract, and are comfortable judging a small vendor on a trial rather than a specification. On a single connection it is the cheaper line for almost any spend a mid-market brand runs, the campaign builder is a genuine asset during a launch phase, and the managed tier is a reasonable middle if you want a person involved without hiring one. We would not try to talk that seller out of it.

Look elsewhere if you run several marketplaces and the base fee multiplies, if you need auditable rules rather than an inferred model, or if procurement will ask for documents that do not exist.

Dr. PPC is a different shape rather than a better bidder, and the difference is worth stating precisely. Instead of moving bids toward a target you typed in, it reads the whole ad account first and writes a strategy for each product against that brand's real unit economics — then every proposed change arrives with the evidence behind it, a plan for measuring whether it worked, and a trigger for rolling it back. You decide how much runs unattended, from every change waiting on a click through to fully autonomous inside guardrails you agreed. Inventory risk, pricing, launches and creative always come back to a person.

It is $300/month plus 3% of ad spend, capped, month to month, first 30 days free, with Orbit included at no additional charge — one fee for the account, however many marketplaces sit under it. Two honest redirects before you compare: if the leak is stock rather than bids, Dr. Stock covers inventory, fees and supply chain and will return more than any bidding change; if sponsored ads are already efficient and growth has stalled, the next question is reach, and that is Dr. DSP. All of them are products of Full Circle, a full-service Amazon management company with $500M+ in managed revenue across 100+ brands.