A Czech contract manufacturer will spend three weeks evaluating a supplier for a part worth four thousand dollars a year, with a scored matrix and two reference calls. The same company approves a marketing subscription of comparable size in a corridor conversation. This article costs a year of automated search work the way the purchasing department would cost anything else: unit price, quantity, line items, total.
The two decisions get such different treatment because one has a specification and the other does not. A component has a drawing, a tolerance and a quoted unit rate, so it can be compared. A campaign arrives as a monthly figure with no visible line items, so it gets approved on impression or refused on instinct. The pricing is arithmetic, and the arithmetic is short.
Put it through your own evaluation process
You already know how to do this. Every exporter selling capacity into a German or Nordic supply chain has been through a supplier evaluation from the receiving end: a scored questionnaire, a request for a breakdown, a challenge on any line that looks like a bundle. Turn that process round and point it at your own marketing budget and most of the confusion disappears.
Three questions do the work: what the recurring unit is and what it covers, what is variable and what drives the variance, and what the commitment is. Those three answers describe the entire cost structure of an automated campaign.
- The recurring unit is a domain, not a project. Subscription is charged per domain per month. One hostname carrying two language trees is one unit, which turns out to matter a great deal for a bilingual exporter.
- The variable part is sold in slots. Placement add-ons are priced per slot at a fixed unit rate, in defined quantities. That is a line item you can multiply, not an estimate.
- The service level is the thing that changes price most. The gap between the two levels is not a feature list so much as who exercises judgement — software alone, or software plus people.
- Time is the input nobody prices. Expect first measurable movement in four to eight weeks, and on thin export search volumes plan for the far end of that.
Two subscription levels, plainly described
AutoSEO — the machine decides
For a company with nobody in the building who owns search, and no plan to hire one.
- Keyword discovery and prioritisation run without you. Candidates are drawn from Search Console, live result pages and any seed terms you supply, then ordered automatically.
- Link building proceeds unattended. Placements come from a partner network of more than 230,000 websites, selected without a manual approval step.
- Full analytics and on-site suggestions included. Eight Search Console views, six rank-tracking views, generative market research and a live chat assistant, all inside the subscription.
FullSEO — people in the loop
For a supplier whose technical vocabulary is specific enough that a wrong word costs a tender.
- Keywords chosen by hand, with automatic fallback. You approve the phrases describing what you are accredited and equipped to do; anything you do not reach is picked up automatically.
- Placement against a domain-rating target. A quality threshold is set in advance instead of taking donors as they come.
- Human review before on-site changes go live. A team of search specialists, developers and writers works alongside the automation, which is the part that stops a page claiming a scope you do not hold.
Over a year on a single domain those two lines are 1,788 USD and 6,000 USD. The difference is 351 USD a month, which is 4,212 USD across twelve months. Whether that gap is worth paying depends on one question, and it is not a question about features: how expensive is it when an automated system describes your capability slightly wrongly on a page a buyer reads before contacting you.
What the extra 351 dollars a month buys
Both levels run on the same unified campaign platform, with the same data connections and the same reporting. Nothing in the analytics is withheld from the cheaper level. What differs is who holds the pen.
| Decision | AutoSEO | FullSEO | Why a supplier might care |
|---|---|---|---|
| Which phrases get targeted | Chosen and ranked automatically | Chosen manually, automatic fallback behind | A process name you are not certified for is a wasted quarter |
| Which sites carry your links | Selected from the network without review | Selected against a domain-rating target | Placement quality is visible to anyone auditing you |
| On-site changes | Suggested, applied on your judgement | Reviewed by people before going live | Capability wording is a claim, not copy |
| Analytics and market research | Complete | Complete | No reason to upgrade for reporting alone |
| Who does the writing | You, with suggestions | Specialists, developers and writers | The scarce resource in most technical firms |
Read that last row as the real pricing question. In a Czech engineering firm or laboratory, the person who genuinely knows what you can and cannot do is a production or quality manager with no spare hours. The higher level is, in practice, a way of buying writing capacity from people who will check the detail with you rather than guess.
Slots, and what a slot actually is
Beyond the subscription, placements are sold as slots at a fixed unit rate in defined quantities — the part that behaves most like ordinary purchasing, and the part most often misread.
| Add-on | Unit rate | Quantities offered | Cost at each quantity |
|---|---|---|---|
| Wikipedia placements | 10 USD per slot | 0 / 1 / 5 / 10 | 0 / 10 / 50 / 100 USD |
| PBN placements | 1 USD per slot | 0 / 20 / 100 / 500 | 0 / 20 / 100 / 500 USD |
The multiplication is trivial and the temptation is obvious. Five hundred slots for five hundred dollars looks like the best unit economics on the page, and a purchasing instinct trained on volume discounts reaches for it automatically. That instinct is wrong here, and the reason is worth stating without any hedging.
Those last two tiles are the envelope. On one domain the cheapest possible month is 149 USD, the subscription alone. The most expensive is the higher level with both add-ons at their top quantity: 500 plus 100 plus 500, which is 1,100 USD. Across twelve months the envelope runs from 1,788 USD to 13,200 USD, and every real budget sits somewhere between.
A year costed in full
Here is a twelve-month plan for a single domain belonging to a Czech exporter with an English capability section. It is staged deliberately: cheap while the data is thin, heavier once there is something to act on.
| Phase | Months | Subscription | Add-ons per month | Monthly total | Line total |
|---|---|---|---|---|---|
| Ramp | 1–3 | AutoSEO, 149 USD | none | 149 USD | 3 × 149 = 447 USD |
| Build | 4–6 | AutoSEO, 149 USD | 20 PBN slots, 20 USD | 169 USD | 3 × 169 = 507 USD |
| Push | 7–12 | FullSEO, 500 USD | 100 PBN slots and 1 Wikipedia slot, 110 USD | 610 USD | 6 × 610 = 3,660 USD |
| Total | 12 months | 3,894 USD in subscriptions | 720 USD in slots | 384.50 USD average | 4,614 USD |
Check it from the other direction, which is what a finance department will do. Subscriptions: six months at 149 USD is 894 USD, six months at 500 USD is 3,000 USD, giving 3,894 USD. Add-ons: PBN slots run 3 × 20 plus 6 × 100, which is 60 plus 600, or 660 slots at a dollar each; Wikipedia slots run 6 × 1 at ten dollars, or 60 USD. Add-ons therefore total 720 USD. And 3,894 plus 720 is 4,614 USD, matching the phase totals of 447 plus 507 plus 3,660.
The comparison worth putting beside it is the flat alternative. Twelve months of the lower level alone is 1,788 USD; twelve months of the higher level alone is 6,000 USD; the staged plan above lands at 4,614 USD, between the two, because it spends the cheaper rate while nothing is known and the higher rate once there is something to steer.
Per domain, not per language
This is the detail that changes the sum for a Czech company more than any other, and it is easy to miss. Charging is per domain. A single site carrying a Czech tree and an English tree is one domain and one subscription, however differently the two halves read.
That is a real saving and a trap in the making. The two trees are not translations of each other; they address audiences with almost no shared vocabulary. The Czech pages are read by people who already know the firm, the English ones by a buyer comparing you against suppliers in Poland, Portugal and Vietnam. One subscription covers both, but only if somebody splits the reporting deliberately.
Czech and English on the same host
A .cz site with an English section under a path. One subscription, one keyword pool, two audiences that must be filtered apart in every report.
- 149 or 500 USD a month covers both trees
- Blended averages hide the export half completely
A separate export site
A .com or .eu property built for foreign buyers alongside the domestic .cz. Two subscriptions: at the lower level 298 USD a month, or 3,576 USD across a year.
- Cleaner separation, double the recurring cost
- Site tags filter the account across both
If you run several properties, the account is built for it. Sites can be grouped under linked Google account groups, released individually to another email address so an external agency sees one property and nothing else, and filtered globally with site tags. The reporting and export layer then produces CSV or JSON at up to 10,000 rows and PDF at up to 250 rows, branded with your own logo and colours — which is how the number in this article becomes a page in a quarterly review rather than a screenshot.
What you are really deciding between
Every purchasing department knows the make-or-buy calculation, and almost nobody applies it to search. The alternative to a subscription is not zero. It is somebody's time, and time in a Czech technical firm is the constrained resource, not cash.
Hours, not salary
The scarce figure is not what a specialist earns. It is how many hours a month a quality or production manager can give to English capability pages, and it is fewer than any plan assumes.
- Count the hours before the cash
- Nobody outside knows your certified scope
Tooling bought separately
Rank tracking, market research, indexing submission and reporting are several line items when purchased apart. Here they sit inside the subscription rather than beside it.
- One recurring line instead of five
- Same data layer at either level
Placement capacity
A network of donor sites is not something a manufacturer builds. It is bought or done without, and that half of the decision is genuinely binary.
- More than 230,000 partner sites
- Quality targeting only at the higher level
Stopping halfway
Search work compounds slowly. Cancelling in month five and restarting in month nine pays for the ramp twice and keeps none of it.
- Budget the whole period or none
- Review at month six, do not restart
That last card turns more reasonable budgets into wasted ones than any other mistake here. First measurable movement typically appears after four to eight weeks, and on the low volumes sourcing phrases carry, expect the slow end of that range. A budget approved for three months is not a smaller version of a twelve-month budget; it is a worse decision, and better not taken.
Whichever level you run, the Stream assistant in the campaign section keeps the record: automatic reports, newly placed links with the donor's rating and traffic, to-do items with states for active, deferred and discarded, and campaign news in one chronological, searchable line. For a spend you will have to justify at renewal, that log is worth more than the dashboard, because it is the only place the sequence of decisions is written down.
Common questions
Can we start at the lower level and move up later?
Yes, and for most exporters that is the sensible order. The first months let the keyword pool fill from real query data, and manual selection is far easier once you can see which terms the system found and which describe work you are certified to do.
Are the add-on slots a monthly commitment?
They are chosen per campaign at the quantities offered — 0, 1, 5 or 10 for Wikipedia and 0, 20, 100 or 500 for PBN — and zero is a valid choice at both. In the worked example above the quantity changes twice across the year, which is the normal pattern rather than an exception.
Our English section sits on the same domain as the Czech site. Do we pay twice?
No. Billing follows the domain, so one hostname with two language trees is one subscription. The work of separating the two audiences in reporting is yours: filter the pages view to the English directory, or use site tags if several properties are involved.
How do we judge whether it worked, given how little volume our phrases have?
Not on clicks, which are too sparse to read on a sourcing vocabulary. Judge on whether phrases entered the top three, ten or thirty bands, on impressions from the countries you sell into, and on enquiries that name a capability. The keyword dynamics view tracks those band transitions directly.
Is there a cheaper route if we only want the analytics?
The analytics are not sold apart from the campaign — the Search Console, rank-tracking and market-research views come with either level. If reporting is all you need, the lower level is the whole answer. Related reading sits in the English articles section.
Building your own version of the table
Take the structure rather than the figures. Decide how many domains you are running — one bilingual site is one, an export property alongside it makes two. Decide the month you would upgrade, which for most exporters is the month the keyword pool holds enough real phrases to select by hand. Decide whether you are buying slots at all, and if so start at the low quantity.
Then multiply. Subscription rate times months, plus slot rate times quantity times months, and check the total twice from both directions the way the worked example does above. What comes out is a figure your finance department can treat like any other recurring supplier line, with a review point written into it. Where a change needs a developer rather than an editor, that work sits with technical SEO and should be costed separately.
To build the table against the real state of your own property rather than an assumption, open the dashboard and connect a verified domain, then read how many English pages currently earn impressions before you decide which level to buy. On most Czech supplier sites that number is smaller than anyone expects, and it changes which phase of the plan you should actually start in.