Two tiers, two add-ons, a slot-based price list — and one step in the middle that decides whether any of it works. For a technical business in Greater Boston, the costly mistake is never the subscription.
Published prices are easy to compare and tell you very little. AutoSEO is $149 per month per domain, FullSEO $500 per month per domain; Wikipedia placements cost $10 per slot and PBN placements $1 per slot. Everything below is arithmetic on those four numbers, plus what the Semalt campaign panel does in return.
The part worth your attention is not the invoice. It is the keyword approval flow: in a market of assays, instruments, protocols and regulated services, an automated pool will propose terminology no specialist would type — and somebody has to be qualified to say no.
The same platform underneath, a different distribution of decisions
Neither tier is a reduced edition of the other. Analytics, rank tracking, indexing tools, generative market research and the Stream assistant are identical on both, reading the same account data. The variable is authority: who makes which decision.
At $149 the platform holds all three levers. It locates and ranks keyword candidates, it places links, it drafts on-site edits, and none of that pauses because nobody signed in this week. At $500 each lever becomes a switch, and the switches are thrown separately.
The words "per domain" carry more weight than they look. A firm with one website pays once. A hospital department with a separate research site, a manufacturer with an acquired microsite, a practice with a legacy domain it never retired — each pays per property, and the properties need not share a tier. For institutional operators that is the most useful lever in the price list, and the most overlooked.
AutoSEO, and what $149 a month is buying
AutoSEO — nobody has to be watching
For a property nobody can supervise weekly, where no single decision justifies a meeting.
- Candidate terms surface on their own. Discovery and prioritization happen before you arrive, so nobody has to assemble a starting list.
- Link placement is delegated entirely. Placements are made across a partner network of over 230,000 sites, with the choosing done for you.
- Page-level edit proposals. Titles, structure and body text drafted against the pages you actually have, not a generic checklist.
- The full analytics stack. Search Console and SERP views alike: clicks, impressions, threshold crossings, rival domains, portfolio-wide position.
- Stream, answering live. A project assistant tied to real campaign data that will take keyword and URL lists in bulk.
What is absent at this price is the veto. The pool stays visible, but the arrangement assumes you watch results rather than approve inputs. For a property nobody intends to open more than monthly, that is usually the right bargain.
It becomes the wrong bargain at a recognizable moment: when a proposed keyword could be professionally wrong rather than merely unproductive. A term naming an indication you are not cleared for, or a method you do not run, is not a small waste of budget here.
FullSEO, and what the extra $351 actually unlocks
FullSEO — the same machine, with hands on it
For regulated wording, vocabulary that has to be exactly right, and teams that already have an editorial plan.
- Hand-picked terms, with a safety net. Select keywords yourself; anything left untouched reverts to the automatic path, so an unattended week never freezes the campaign.
- Placements signed off one by one, against a rating you set. Name the domain rating you are targeting and approve each placement rather than inheriting the automatic pick.
- On-site changes held for a person to read. Edits wait in a queue instead of publishing themselves — the control that decides whether counsel or a compliance officer can live with this at all.
- People, not only software. The tier carries SEO specialists, developers and writers working alongside the campaign automation.
Those switches are independent, and mixing them is the normal case. A diagnostics company may choose its own terms because it knows precisely which assays it sells, leave placements automatic because nobody internally holds a view on donor sites, and still require that every line of patient-facing text be read before it goes live.
Where the keywords come from, and who is qualified to reject one
The keyword pool draws on three sources at once, which is why it never resembles the list you would have written unaided.
What is already measurable
Search Console contributes terms you demonstrably appear for, several of which nobody chose. Live SERP data contributes the shape of the results page for those terms today.
- Real impressions, not volume estimates
- The current competitive picture
- Reporting lag already accounted for
What only your people know
Your own seed keywords: the platform you just validated, the sample type you now accept, the term your field application scientists hear on every call.
- Entered directly or batched through Stream
- Covers terms with no search history yet
- Where internal expertise enters the system
Candidates then arrive one at a time for a verdict: in, out, or later. The third state is the underused one and the most valuable here. A term tied to a validation finishing in spring, or a service you are still hiring for, is not permanently wrong — parking it withholds it without discarding the research.
This is where a generic pool meets a specialist market. The first two sources report the open web honestly, and the open web phrases these subjects like a consumer, so the pool offers broad and comfortable terms. The revenue in Greater Boston attaches to vocabulary a consumer phrasing never contains.
| What a generic pool tends to propose | What the buyer actually types | Why the difference matters |
|---|---|---|
| Broad service category plus city | The method, platform or assay by name | The broad term draws students and job seekers, not purchasers |
| A layperson's name for a procedure | The clinical or regulatory term for it | Different terms describe different things; one of them is wrong |
| Popular adjacent technology | The specific instrument and configuration in use | Ranking for a technology you do not run wastes every click |
| A general "solutions" or "services" phrase | Sample type, throughput, turnaround, tolerance | Specific constraints are how a qualified buyer shortlists |
None of that is a defect in the automation. It is the automation faithfully reporting a web written mostly by non-specialists, and correcting it is the reviewer's job — the part of this arrangement organizations get wrong.
The task usually lands on whoever owns the marketing calendar, and that is frequently the wrong person, because three of the judgments required are technical. Whether two terms name the same thing. Whether a phrase implies a claim you are not permitted to make. Whether the people typing a term could ever buy anything, or are undergraduates writing a literature review.
- Name one reviewer, not a committee. A single person with domain knowledge and the authority to reject. Committees produce backlogs, and a backlog lets the automatic fallback decide for you.
- Twenty minutes a week is enough. The pool arrives in batches. Treating this as a project rather than a habit is why it gets postponed until it stops happening.
- Reject with a one-line reason. "Different assay", "not cleared for that", "academic traffic only". Three months of those lines is a usable brief for whoever writes next.
- Defer anything tied to a date. Pending validations, clearances, hires and launches. Deferral holds the research until the term becomes true.
- Feed the reviewer's own vocabulary back in. Seed terms and URL lists can be handed to the Stream assistant in batches, so the correction becomes an input rather than a monthly argument.
What the two placement add-ons cost
Neither add-on is an open budget. Both sell as fixed blocks of slots, which makes them easy to forecast and hard to overspend on unintentionally.
Wikipedia placements
Four block sizes — none, one, five or ten — billed by the slot.
- The whole thing costs less than a week of the base fee. Ten slots total $100, the smallest line in any realistic budget.
- A choice from four, not a dial. No sliding scale to agonize over: pick one block size and move on.
PBN placements
Four block sizes — none, twenty, one hundred or five hundred — billed by the slot.
- Picking the block size is the entire decision. Twenty slots costs $20 monthly, one hundred costs $100, five hundred costs $500 — the largest block matches the whole FullSEO fee.
- The invoice scales; nothing else is promised to. Five times as many slots is reliably five times the money and reliably nothing else.
| What is billed | Charged by | Rate | Choices | Ceiling per month |
|---|---|---|---|---|
| AutoSEO subscription | Domain | $149 | One tier | $149 each |
| FullSEO subscription | Domain | $500 | One tier | $500 each |
| Wikipedia placements | Slot | $10 | 0 · 1 · 5 · 10 | $100 |
| PBN placements | Slot | $1 | 0 · 20 · 100 · 500 | $500 |
A year costed for a Cambridge diagnostics company
Take a mid-sized specialty diagnostics business in Cambridge selling to hospital laboratories and biotech. It runs a corporate domain, and it also owns a product-support microsite inherited with an acquisition, still ranking for instrument and consumable queries nobody has thought about since the deal closed.
The corporate domain carries the regulated claims, so it takes FullSEO with human review switched on. The microsite is technical, low-risk and unattended, so it takes AutoSEO — but not until month four, once someone has confirmed which of its pages are worth keeping. Both sit in one account, separated by site tags.
| Subscription or add-on | Which property | Active in months | Per month | Year |
|---|---|---|---|---|
| FullSEO | Corporate domain | 1–12 | $500 | $6,000 |
| AutoSEO | Product microsite | 4–12 | $149 | $1,341 |
| Wikipedia, 5 slots | Corporate domain | 6 and 7 | $50 | $100 |
| PBN, 20 slots | Product microsite | 7–12 | $20 | $120 |
| PBN, no slots | Corporate domain | never | $0 | $0 |
| Total for the year | $7,561 |
The arithmetic closes as follows. Three months at $500 is $1,500; months four and five at $649 add $1,298; month six at $699 and month seven at $719 add $1,418; the last five months at $669 add $3,345. Total $7,561, an average of $630 a month, with add-ons accounting for $220 — under three percent.
Three choices in that schedule survive any change to the numbers. Subscriptions precede add-ons, because volume purchased in month one arrives before anybody has approved a keyword. The second property waits until a person has decided what it is for. And the regulated domain carries no PBN slots whatsoever — an explicit zero rather than an omission.
A realistic schedule, and the tier each situation calls for
Most dissatisfaction with automated campaigns turns out to be a calendar problem. Reckon on four to eight weeks before anything measurable moves, a quarter or two before a trend can be defended, and a full year before the picture resolves — which is why the schedule above was drawn over twelve months.
Locally there is a second clock. Search behavior across Greater Boston turns over with the academic year, so a campaign begun in June spends its first weeks in the quietest stretch of the calendar and looks flat for reasons unrelated to the work. Judge it against the same months a year earlier, not against the spring you just left.
Where $149 is the right answer
Properties that nobody will supervise and whose wording carries no regulatory weight.
- An acquired or legacy microsite
- A departmental or campus subsite
- A second brand kept for its history
Where $500 earns its difference
Properties where a wrong word is a problem rather than a wasted click.
- Clinical and diagnostic claims
- Financial disclosure language
- Specification-level terminology
The reviewer, not the budget
The upper tier is worth its price only if a qualified person actually uses the controls.
- Twenty minutes weekly, named person
- Otherwise the switches idle
- Then the cheaper tier is honest
Add-ons last, sometimes never
Placement volume belongs after the pool is settled and the page work has begun.
- Never in the opening month
- Zero on a name-sensitive domain
- Reviewed again each quarter
Questions that come up
How soon should we expect to see anything?
Rankings and impressions typically begin shifting somewhere in the four-to-eight-week band. Signed contracts do not: in sectors with long procurement cycles the search metrics move months ahead of anything reaching the pipeline. Allow two quarters before passing judgment.
Can we start on AutoSEO and move up later?
Since each property carries its own tier, nothing structural stands in the way. The signal to move is not a bigger budget but the arrival of opinions: once you care which terms are pursued, which donor sites are acceptable, or who reads a page before it publishes, the entry tier has stopped fitting.
Who should approve keywords if our marketing team is not technical?
Someone technical, with twenty minutes a week and the authority to say no: a product manager, a senior scientist, a practice lead. Marketing can prepare and summarize the batch; the accept-or-reject call needs a person who knows whether two terms name the same thing and whether a phrase implies a claim you cannot support.
Should we buy the placement add-ons at all?
Wikipedia slots are small money and easy to trial: ten of them come to $100. PBN slots need a decision. Twenty or a hundred is cheap enough to test on a low-risk property; five hundred costs what the upper subscription does and should never be anyone's default. On a regulated or heavily branded property, none is a legitimate answer.
We have four domains from old projects. Do they all need subscriptions?
Decide what each one is for before pricing anything. Several should probably be redirected into the main site, which costs nothing monthly. The ones holding genuine standing on technical queries are candidates at $149. Paying a subscription on a domain nobody intends to maintain is the most common avoidable line in an institutional budget.
The rates are published; the size of the bill is your decision
The arithmetic is the easy half. Rates do not move and anyone can build the table. What moves the figure at the bottom is a short set of choices: how many properties you genuinely intend to maintain, what tier each deserves, and whether the add-ons were decided or merely defaulted.
The choice that changes the outcome rather than the total is the one this article kept returning to. An automated campaign is a machine for executing decisions at volume, and in a market where the vocabulary is technical and half the terms mean something specific, the quality of what it executes is set at the approval step. A qualified reviewer for twenty minutes a week is worth more than the difference between the two tiers. Our service overview describes how that review runs alongside a client campaign, and the blog archive covers the analytics that show whether it is working.
The tiers, the slot blocks and the campaign module against a live property can all be inspected directly — sign in and set up your first domain. Then, before the first invoice clears, settle who does the rejecting. That name matters more than the tier.
