The price list holds two numbers, 149 and 500 dollars a month. Neither tells you what a tier will cost you: that is decided by how many domains you run and by the hours someone in-house spends on decisions.
In an exporting manufacturer both multipliers run higher than you would guess: one product, several territories, each on its own domain in its own language.
Finnish volumes are small, the German-speaking market densely contested, and Swedish results mix in wherever product names are shared.
Semalt's campaign automation splits into two tiers, with two kinds of link placement sold on top. Start with what the invoice follows — that is where a multi-market budget breaks.
Four figures and one billing unit
The whole structure fits into four numbers: two monthly subscriptions and two per-unit add-ons in fixed steps. No intermediate tiers, no bundles, no discounts — easy to compare, impossible to negotiate.
| Item | Price | Invoice follows | Steps |
|---|---|---|---|
| AutoSEO subscription | 149 USD / month | One domain | None |
| FullSEO subscription | 500 USD / month | One domain | None |
| Wikipedia placements | 10 USD each | Placements bought | 0, 1, 5 or 10 |
| PBN placements | 1 USD each | Placements bought | 0, 20, 100 or 500 |
At that rate the base tier lands near 134 euros a month and the guided tier near 450, per domain, excluding VAT — but the total misleads while your own hours are missing.
AutoSEO: a campaign that waits for nobody to log in
The base tier is built for the case where search optimisation is nobody's main job. In manufacturing that is the rule: the marketing person also runs the fair stand, the catalogues and the dealer newsletter.
AutoSEO — the machinery covers your fair weeks
For companies where optimisation is one person's side duty and results are still expected.
- Keywords are found and ranked without you. Candidates are gathered, duplicates stripped, the list ordered by priority; what is left is a verdict.
- Links accumulate steadily. Placements are built into the partner network continuously, each entry logged with the host site's metrics.
- On-site changes arrive pre-drafted. The AI proposes titles, descriptions and internal references; you approve or reject.
- The full instrument panel is included. Analytics, rank tracking, competitor views and report exports belong to the base tier — nothing is reserved for the dearer subscription.
- Stream answers with real figures. The project's AI assistant reads the campaign's own data and takes keyword and URL lists in batches.
A year on one address costs roughly the rent on a modest fair stand — but the stand is gone in three days and the campaign runs twelve months.
FullSEO: the same gears, three points for a human hand
The upper tier does not replace the lower one, it contains it whole. On top come three points where an automatic decision becomes a human one, so part of the price is working hours.
FullSEO — manual selection and pre-publication review
For markets where the biggest volume and the best margin sit on different keywords.
- Keywords are picked by hand. You choose which candidates go into production; if a month passes without a pick, automation continues as a fallback.
- Links are ordered at an authority level. You set a Domain Rating target for host sites rather than accept what the network offers.
- Changes wait for approval. Human-review mode holds on-site changes for inspection — a necessity on pages carrying dimension tables and certificate references.
- There is a team behind it. Specialists, developers and writers come with the tier, needed where software stops helping: German copy that must be rewritten, not translated.
- The mode switch is always available. You move between manual selection and automation at any time; the choice is not locked when you subscribe.
That switch is the structure's most practical detail and still goes unnoticed. Early on manual selection has no basis, so automation collects; with two quarters of data the picks land far more accurately.
The store both tiers draw on
Whichever tier you pick, the campaign works from the same keyword store. Three sources fill it, and in a small language market none alone covers what customers type.
Google Search Console
Actual impressions and clicks from your own address, nothing estimated.
- Shows positions from page two down
- Blind to searches where you never appeared
Live SERP
The results page as it stands now, filling in where your own history cannot reach.
- Reveals German and Swedish players in the topic
- Shows how crowded the top is before you buy
Your own seed terms
Trade language, abbreviations and standards used only inside the industry — the one source you control.
- Most often left unfilled
- Affects the outcome most
Languages crossing
One Finnish concept splits into many written forms, and three languages compete over shared product names.
- Case endings and compounds multiply rows
- Swedish results blend in under shared names
The third card is most often left blank, and in manufacturing that is expensive. The tool does not know that customers ask by a trade name in no public dataset, or that tenders cite a standard number and never the product.
The fourth explains why the length of a candidate list deceives: ten rows may mean three real topics in different case forms, or ten separate topics. Volume deceives too: forty monthly searches can beat a thousand-search general word, if those forty come from a procurement manager tendering a year's supply.
Every candidate passes an approval step, logged with user and date. That happens in the campaign's candidate view, identical on both tiers.
- Approved. The word matches what you sell, the page exists and you can deliver — including whether you ship to that country at all.
- Rejected. Consumer searches are pure traffic to an industrial supplier, even at ten times the volume of the commercial terms.
- Queued. The word is right, but the page does not exist or is not translated. Without it every round would discard the terms whose value shows later.
Wikipedia and PBN: same product group, opposite logic
Two kinds of link placement sit on top of a subscription, on identical terms at either tier, and zero is valid in both. The tenfold price gap measures scarcity, not effect.
Wikipedia placements
Few placements, a high unit price and a volunteer editorial body assessing every addition.
- Four steps, a low ceiling. Zero, one, five or ten placements; the top step is a hundred dollars a month, roughly 90 euros.
- Justifiability decides. Only what evidences a claim stays in an article — a patent, standards work or local history, or nothing.
- Permanence is the whole deal. If the addition is removed within a week you paid for nothing, and the risk is yours.
The second add-on is structurally the opposite, and its low unit price is exactly why it deserves the longest pause.
PBN placements
Many links for little money through a site network — of the four prices, the one needing longest thought.
- The steps climb high. Zero, twenty, one hundred and five hundred; the top rung costs five hundred dollars a month, as much as an entire guided campaign.
- What you buy is a quantity. Ten encyclopaedia placements cost the same as a hundred network ones, which describes what a single placement weighs.
- The name carries the warning. A private blog network exists to carry links; cheapness and unlimited replicability are its selling points and its marks.
Add-ons and tier are separate decisions: placements sell in the same steps whichever tier you pay, and the risk profile differs, as often does the decision-maker.
A timber products plant opens up a year
| Item | Option A — four addresses, base tier | Option B — German address guided | Basis |
|---|---|---|---|
| Subscriptions, 12 months | 7,152 USD | 11,364 USD | four × 149 versus 500 plus three × 149 |
| Wikipedia placements | 0 USD | 120 USD | none versus one on the German address |
| PBN placements | 0 USD | 0 USD | zero in both, as argued above |
| Total invoiced | 7,152 USD (approx. 6,440 €) | 11,484 USD (approx. 10,340 €) | excluding VAT, currency risk uncounted |
| Internal working time | 48 hours = 3,552 € | 96 hours = 7,104 € | 74 € per hour, four addresses |
| Pages and translations | 10 pages = 4,500 € | 22 pages = 9,900 € | 450 € per page, four languages |
| Outside the invoice | 8,052 € | 17,004 € | hours and copywriting combined |
Two things stand out. The subscription fee is a minority of total spend: in option B a good third reaches the invoice, the rest hours and writing nobody books to marketing.
The second is the asymmetry in option B. The German-speaking area is the most crowded and the deals largest, so it benefits from manual selection; the Swedish address, on a few enquiries a month, does not. Raising all four to the same tier is the most expensive way to avoid taking a position.
The revenue side is left out deliberately: in industry the sales cycle often runs six months to two years, so first-year margin says almost nothing. The cost side is plannable, and it is the half you can influence.
A realistic schedule, and what gets measured before it
An automated campaign tempts you into checking results every morning, and that is wasted time. Some tracked quantities move in weeks, others only after six months.
- Month one — signs of life. One question matters: is the machinery moving. Are keywords processed, placements appearing, proposals arriving that need a verdict.
- Months two and three — first indications. Impressions grow on processed words and back-row positions climb a few places; clicks follow later.
- Months four to six — visible change. Organic traffic grows and the searches that matter to sales reach page one, the domestic address usually moving first.
- Late in the year — enquiries. Contacts, requests for quotation and dealer questions. Only at this point is the spending weighed.
The interval is easiest to follow through the Stream feed, where new placements, reports and tasks run in one chronological stream. It speeds nothing up, but keeps a quiet month distinguishable from a stopped one.
If nothing moves after two quarters, the reason is rarely the tier: more often thin content, a technical foundation putting up resistance, or a market where the budget does not stretch. The first two can be checked in advance, alongside our technical services.
Who each tier suits
Subcontract machine shop, two languages
A few large accounts, long contracts, a narrow vocabulary; precision beats volume.
- Base tier on both addresses
- Seed terms from tender documents
- No add-on placements
Food producer, strong at home
Domestic demand solid, export at trial stage, no data yet for manual selection.
- Automation collects for two quarters
- Reassess control at six months
Timber industry, four addresses
Competitive density and deal size vary by market, so a uniform tier overpays on one and underserves another.
- Control on the heaviest market
- The others stay on the base tier
- Reviewed twice a year
Seasonal business in the regions
Demand concentrates into a few months, but the searches happen months earlier.
- The subscription runs all year
- Manual selection before the booking window
- Automation after the season
| Area | AutoSEO | FullSEO |
|---|---|---|
| Monthly price per address | 149 USD | 500 USD |
| Keyword selection | Automatic discovery and prioritisation | Manual, automation as fallback |
| Link placements | Continuous stream from the network | Also targeted to a chosen DR level |
| On-site changes | As AI proposals | Human-review before publication |
| Who does the work | Software and the Stream assistant | Also specialists, developers and writers |
| Analytics and reporting | In full | In full |
| Add-on placement steps | The same | The same |
| Switching mode mid-term | Not applicable | At any time |
Those rows describe capabilities; your choice comes down to three questions. Is there a named person with four to six hours a month for picks and approvals? Are the most-searched terms and the best-margin products different? Do changes need checking before publication, because pages carry dimensions and certificates? Two yes answers argue for the guided tier, and the figures sit in the rank history and competitor views.
Common questions
Can add-on placements be bought without either subscription?
Placements are chosen on top of a subscription, so the campaign tier comes first. Steps and unit prices are the same whichever tier you pay, so a base-tier customer misses nothing on the add-on side.
What happens if nobody picks keywords one month?
On the guided tier automation continues as a fallback, so the campaign does not stall. That matters for an exporter, where fair season, the year-end and the holidays each take marketing's attention in turn. If three empty months accumulate, ask whether you are paying for control you do not use.
A hundred dollars a month on placements — where is it best spent?
The same sum buys ten encyclopaedia placements or a hundred network ones. The first needs a justifiable role — a patent, standards work, local history — and without it the money burns; the second needs nothing, which is why its benefit is least certain. A third option is one properly written product page.
Is the same data visible on both tiers?
Yes. Analytics, rank tracking, keyword dynamics, competitor views and report exports belong to the base tier in full. The difference is who makes the choices and checks the changes — worth verifying when comparing suppliers, since on many price lists the panel opens only at the top tier.
How is seasonal demand handled?
With the mode switch: time manual selection to the weeks before a season or a major fair, then return to automation. Searches happen well before the purchase, so tightening during the season is already late. More scheduling examples are on our blog.
The question is not two price tags but the way of working your house can commit to for twelve months. The lower tier sells a process that turns by itself; the upper sells influence, wasted without a place in the calendar. A budget calculation works only once internal hours sit beside the subscription fee.
To test the split on your own markets, open the panel and compare the tiers with your own domain's figures and see how many proposed keywords you would let through in each language. That number indicates the tier better than any feature comparison.