In a token model you pay a monthly fee that gives you a bank of credit, and every deliverable in the catalog has a fixed, published price: a reel costs 40 tokens, an SEO audit 100, a full landing page 280. You request what you need that month and the balance is deducted. It makes sense when your volume of work is real but variable, and it does not when you always produce the same thing or nobody on your team can decide what to ask for.
Hiring digital marketing usually means choosing between two extremes: an agency retainer with a fixed scope, or freelancers per project with no continuity. The token model is a third way that has been working for years in the United States, where platforms like Design Pickle and Superside educated the market, and that is only now starting to appear in Latin America.
What exactly is a token model
It is a change in the unit you buy. On a retainer you buy time: a team's capacity for a month. In a token model you buy outputs: a bank of credit you exchange for concrete deliverables with a price known in advance.
The practical consequence is that the price stops depending on how many hours someone took to make your piece. A reel costs the same whether the team produced it in two hours or six. That risk moves from the client to the supplier, which is exactly where it should sit.
How it works, step by step
- You choose a monthly plan. Each plan includes a fixed number of tokens: 1,000 on the Boost plan, 3,000 on Scale, 8,000 on Agency.
- You request what you need from the portal. When you pick the service you see its cost in tokens before confirming, and the tokens are put on hold.
- The team accepts the request and that is when the balance is deducted. If for any reason it cannot be taken on, the tokens return to your bank in full.
- You receive the deliverable with performance data, and that information shapes what you request the following month.
What each thing costs
This is the most appreciated part of the model: prices are public and the same for every client. A sample of the current catalog:
| Deliverable | Cost in tokens |
|---|---|
| Social media post (design and copy) | 30 |
| Reel or TikTok up to 60 seconds | 40 |
| Blog article up to 800 words | 50 |
| Monthly content calendar | 80 |
| SEO audit | 100 |
| On-page optimization for up to 5 pages | 120 |
| Monthly paid media campaign management | 200 |
| Full landing page | 280 |
| Website of up to 5 pages | 400 |
What a real month with 1,000 tokens looks like
Numbers in the abstract say little. This is how the bank of the smallest monthly subscription plan splits across a normal month of operation:
| What was requested | Quantity | Tokens |
|---|---|---|
| Paid media campaign management | 1 | 200 |
| Social media posts | 8 | 240 |
| Reels | 4 | 160 |
| Blog articles | 2 | 100 |
| SEO audit | 1 | 100 |
| Content calendar | 1 | 80 |
| Automated email sequence | 1 | 80 |
| Total consumed from the monthly bank | 18 deliverables | 960 |
What happens to the tokens you don't use
This is the most important difference compared with a retainer, and also with most design subscriptions. Every token is valid for 60 days from the moment it enters your bank. If you paid your subscription in August and only used half, the other half adds to September's tokens.
In practice that means a slow month is not money lost, but capacity saved for the next one. Only what is still unused when the 60 days are up expires, so you always have up to two months of tokens available to concentrate on a launch or a high season.
How it differs from unlimited design
Unlimited design subscriptions promise unlimited requests, but handle them one at a time and in order of arrival. The limit exists: it is the queue. In a token model the limit is explicit and you see it before committing, which for planning purposes is considerably more useful than an unlimited offer with fine print.
The other difference is scope. Unlimited design delivers graphic pieces. A token model applied to full marketing also covers paid media, organic search, content, social and web development, which is where the work starts to connect with business results rather than just good-looking deliverables.
When it makes sense, and when it doesn't
It makes sense when your volume of work is real but irregular: launch months with heavy production of pieces, maintenance months focused on search and measurement. In that scenario a fixed retainer wastes budget and hiring separate freelancers loses coordination across channels.
It does not make sense in two cases. If your production is high and constant, identical month after month, a well-negotiated retainer will come out cheaper per deliverable. And if nobody in your company has the time or the criteria to decide what to request each month, the token bank goes unused: the model hands the client the decision about priorities, and that freedom only helps if someone exercises it.
What usually goes wrong with these models
The weak point of most subscription platforms is that they deliver tasks, not impact. They produce the piece, send it and close the ticket. Nobody checks whether it worked, and the following month the cycle repeats with the same blind spot.
It is a structural problem: if you charge per deliverable, your incentive is to produce deliverables. That is why, before signing up to any consumption model, it is worth checking what happens after delivery. If the answer is nothing, you are buying production volume, not growth.
Frequently asked questions about the token model
Is a token the same as an hour of work?
No. The token is a unit of price, not of time. Each deliverable in the catalog has a fixed cost based on the work involved, and that cost does not change if the team takes more or less time to produce it. You pay for the result, not the hours.
Can I request something that is not in the catalog?
Yes, it is quoted as a special project: you are told how many tokens and how many business days before you decide. The deadline starts counting when you accept the quote, not when it is sent to you.
How many things can I request at once?
It depends on the plan: two active requests on Boost, four on Scale and unlimited on Agency. You can create as many as you like; those beyond the limit wait in the queue and start when a slot frees up.
Can I try the model before subscribing?
Yes. The Explorer pack costs a one-off 10 USD and includes 200 tokens for a real project, with no subscription. If it works, you move to a monthly plan keeping your history and your files.