Pricing & Controlling Costs
Costs on Septenar scale with platform usage — primarily survey responses, with a small set of features that use credits of their own, such as translation, AI analysis and email campaigns, and all of it weighted by the app's compliance level. This guide describes the credit model, plan tiers, the features that use credits, what happens when included credits run out, and the controls available to cap spending.
The credit model
The base unit is one answer. Each question a participant answers uses one credit. A 10-question survey with 50 responses where every question is answered uses 500 base credits. Skipped questions don't count.
Each plan includes a monthly credit allotment. Beyond that allotment, paid plans bill additional usage at a per-tier rate.
| Plan | Monthly credits | Price | Additional usage |
|---|---|---|---|
| Evaluation | 500 | Free | hard cap (no additional usage) |
| Basic | 4,000 | $9/month | $2.00 per 1,000 credits |
| Core | 14,000 | $29/month | $1.80 per 1,000 credits |
| Advanced | 55,000 | $99/month | $1.60 per 1,000 credits |
This is the base credit cost. Translation, AI scoring and email campaigns add to it, and compliance level multiplies the whole. Each is described below.
Plans also include limits beyond credits — number of applications, active forms, team members, participant pool size, available compliance levels. Those are at /plans.
A worked example: base credits only
A 12-question survey on the Core plan ($29/month, 14,000 credits included) collects 2,500 responses. Every participant answers every question.
Total: 2,500 × 12 = 30,000 credits.
Of those 30,000 credits, 14,000 come from the included monthly allotment. The remaining 16,000 are billed as additional usage: 16 × $1.80 = $28.80.
Month total: $29 + $28.80 = $57.80.
This example uses base credits only — no compliance multiplier, no translation surcharge, no AI scoring. Adding any of those changes the credit total, not the per-credit price.
Roll-over
Unused credits accumulate in the account balance and are drawn from before additional-usage billing begins. Credits accumulate up to 60 months of the plan's monthly amount, after which the surplus stops adding.
Example, on the Basic plan (4,000 credits/month):
| Month | Added | Used | Balance |
|---|---|---|---|
| January | 4,000 | 1,000 | 3,000 |
| February | 4,000 | 500 | 6,500 |
| March | 4,000 | 12,000 | 0 (+ 1,500 additional) |
| April | 4,000 | 0 | 4,000 |
In March, the accumulated 10,500 absorbs most of the spike; only 1,500 answers are billed as additional usage.
When credits run out
Behaviour differs by plan.
Evaluation. Submissions stop. Forms display a message indicating they're no longer accepting responses. No charges occur — Evaluation accounts have no payment method on file. To resume collection, wait for the next month's allotment or upgrade to a paid plan.
Basic, Core, Advanced. Submissions continue, and additional usage is billed at the tier's rate, up to the account's spending limit (see Controls). Below the limit, forms do not reject submissions because of billing state; at it, they stop until the next month or until an owner raises it.
Compliance multipliers
Compliance level is set per app and applies to every form within it. Higher levels increase the credit weight of each answer to reflect the additional processing, encryption, and audit logging involved.
| Level | Credit weight | Available on |
|---|---|---|
| PIPEDA Baseline | 1.0× | All plans |
| Law 25 | 1.3× | Core, Advanced |
| HIPAA | 1.5× | Advanced |
A 12-question response on a Law 25 app uses 15.6 base credits (12 × 1.3). The credit cost calculator in Survey Settings → Limits & Cost shows the exact credits per submission for the form's question count and the app's compliance level.
The multiplier applies to everything the app uses credits for: answers, translation, AI scoring, email campaigns, custom-domain serving and theme discovery alike.
Other things that use credits
These optional features add to the credit total beyond base submissions, each weighted by the app's compliance level like an answer.
- Translation surcharge — 0.1 credits per answer when a participant submits in a language other than the form's default. A 12-question response in the non-default language adds 1.2 credits to the base 12 on a PIPEDA Baseline app, and 1.56 to the base 15.6 on a Law 25 app.
- AI-assisted analysis of free text — 0.5 credits per scored response, per text question with AI scoring enabled. Described in the next section.
- Email campaigns — 0.1 credits per email sent, reminders included. A campaign to 5,000 pool members uses 500 credits on a PIPEDA Baseline app, and 650 on a Law 25 app, before any responses arrive.
- Custom-domain serving — 0.2 credits per answer submitted through the app's custom domain.
- Theme discovery — 0.3 credits per response a discovery run reads. Described in Understanding theme discovery.
Each appears as its own line item in the billing dashboard.
AI-assisted analysis of free text
Free-text responses — "What would you change about the program?", "Describe your experience.", "What other supports would have helped?" — are valuable but expensive to read at scale. With a few hundred responses, manual reading is fine. With several thousand, it becomes a multi-week project.
AI-assisted analysis converts free-text answers into structured categories using a rubric the builder defines at design time. The platform reads each response against that rubric and writes back a category, score, or sentiment value that can be tallied and charted alongside the survey's other answers.
The rubric is defined at design time
When AI analysis is enabled on a text question, the builder chooses one of four rubric types:
- Sentiment — categorizes each response as positive, negative, or neutral.
- Agreement — scores each response on how strongly it agrees or disagrees with a stated position the builder provides.
- Quality — scores each response on completeness or thoughtfulness.
- Custom categories — the builder defines the categories. For a question like "What would you change about the program?", the builder might define: cost, scheduling, staff communication, eligibility criteria, facilities, other.
A short instruction the builder writes — for example, "classify the response by what aspect of the program is being criticized" — directs the categorization. The instruction can reference other questions in the same survey using {{q:question-uuid}} syntax, so the AI can take prior answers into account when scoring a free-text response.
Both the rubric type and the instruction are reviewed in the survey's AI Analysis settings before publishing. Once the form is published and responses begin arriving, the rubric is fixed for those responses.
What it costs
0.5 credits per scored response, per text question with AI scoring enabled. The surcharge is added to the base credit cost for that response and, like every use of credits, carries the app's compliance multiplier.
Consider an independent evaluator running a program evaluation for a housing nonprofit. The survey has 12 questions; three are open-ended (what would you change, describe your experience, what other supports would have helped). 2,500 program participants respond. The account is on Core, with PIPEDA Baseline compliance.
| Component | Calculation | Credits |
|---|---|---|
| Base submissions | 2,500 × 12 × 1.0 | 30,000 |
| AI scoring | 2,500 × 3 × 0.5 | 3,750 |
| Total | 33,750 |
Of those 33,750 credits, 14,000 are included on Core. The remaining 19,750 are billed as additional usage: 19.75 × $1.80 = $35.55. Month total: $29 + $35.55 = $64.55.
Without the AI scoring, the same survey costs $29 + $28.80 = $57.80 — 30,000 base credits, 16,000 additional. The AI scoring contribution is $6.75, which categorizes 7,500 free-text answers (2,500 responses × 3 questions) against the rubrics defined at design time.
If the same survey ran on a Law 25 app (Core or Advanced), both would carry the 1.3× multiplier — 39,000 base + 4,875 AI = 43,875 total — and the additional-usage portion grows accordingly.
What it can't do
The rubric is set at design time. If the categories prove wrong after responses arrive — the question elicited themes the rubric didn't anticipate — the rubric can be revised, but every response scored against the original rubric needs to be re-scored against the revised one. Re-scoring incurs the surcharge again.
The AI confirms what the rubric tells it to look for. It will not surface themes the rubric wasn't asked about. A custom-categories rubric of cost, scheduling, staff communication will not flag responses about language access or transportation — those will land in other if defined, or be misclassified if not. Custom rubrics work best when the builder has read 50–100 responses first, drafted the categories from what's there, and used the AI to scale that categorization across the rest.
For exploratory analysis — when the goal is to discover unexpected themes rather than confirm anticipated ones — manual reading is the better approach.
When to use it, when to skip it
AI scoring is appropriate when the response volume is large enough that manual reading is impractical, and the categories are well-defined enough that a rubric can capture them.
Skip it when:
- Responses will number in the low hundreds. Manual reading is faster, more accurate, and surfaces nuance the AI won't.
- The free-text questions are exploratory — designed to discover themes rather than confirm them.
- The rubric is unstable at the time the survey publishes. Revising mid-collection means re-scoring.
- Automated categorization without human review would be inappropriate for the data at hand. The output of AI scoring is a starting point that human analysts review and refine.
Controls
Three ceilings work independently.
Spending limit (per account). The most a paid account can use beyond its remaining credits in a calendar month. It is on by default:
- In the first billing cycle, it is one month's included credits, and it clears when the first renewal is paid. Support can lift it sooner once they have confirmed who holds the account.
- After that, it is the plan's default — five months of included credits at the plan's rate: $40 on Basic, $126 on Core, $440 on Advanced — until an owner sets an amount in dollars or removes it.
An owner can raise, lower or remove the limit on the billing page; an administrator can only lower it. When it is reached, every form in the account stops accepting responses, and campaigns and theme discoveries are refused, until the next month or until an owner raises it. Participants who try to respond are told the form can't accept responses right now, and nothing about the account's billing.
Credit limit (per app). A monthly credit ceiling on one application — an agency's allocation for one client, for instance. When reached, that application's forms stop accepting responses for the rest of the month. Set on the application.
Submission cap (per form). A hard limit on total submissions to a given survey or poll. Once reached, the form stops accepting responses. Useful for research with a target sample size — "200 responses, no more." Set in Survey Settings → Limits & Cost.
A submission cap addresses one form's volume, an app's credit limit one application's share, and the spending limit the account's total exposure. An owner who removes the spending limit and sets no other ceiling has responses continue to accumulate and additional usage continue to bill.
Alerts
Email alerts are sent at 80% and 95% of the month's credits (in the first billing cycle, of the first-cycle limit), and at 80% of the spending limit and when it is reached. They go to the account's billing address and to its owners and administrators. The billing dashboard shows real-time balances, the month's usage against the spending limit, and per-app usage breakdowns.
Configurable billing alarms can be set on a credit-usage threshold or on a dollar threshold for additional usage, at the account or per-app level.
Pausing
Paid accounts can be paused from the billing page, for as long as needed. While paused, nothing is charged, no credits are used, and the accumulated balance is held exactly as it is. Forms stop accepting responses and campaigns will not send; surveys, results, exports, and team access all remain available. The account owner can resume at any time — resuming starts a new billing cycle that day, with the cycle's charge and credits both applied immediately. A pause that ends within a period already paid for simply continues on the original schedule.
The held balance cannot be spent while paused — pausing is a way to stop the meter, not a way to run surveys without a subscription.