Independent editorial guide, written with AI assistance. Examples are hypothetical; no hands-on product testing or owner interviews were performed. Features and prices may change. Our editorial method.
ShiftToAGI, Shifting Founders Beyond AI
Evidence note, checked 2 November 2026. We could retrieve only one official source, the FAQ on Make's pricing page. Make's AI page and the English-language pricing URL returned errors when checked. This profile therefore makes few product claims. Where the retrieved text did not document a capability, we say so. Use this as a planning aid, not a full feature review.
Make is a tool for building what it calls scenarios. Its pricing FAQ says a scenario connects applications, transforms data and sends that data to another app. Usage is billed in credits. Most actions use one credit, and some advanced Make AI Provider features use more (source).
That is the documented core. The rest of this article explains what it could mean for someone who runs a business rather than writes code, and where you need to check before relying on it.
A hypothetical starting point
This story is fictional and illustrative.
Priya runs a small company that sells commercial kitchen equipment through regional distributors. Enquiries arrive through a web form and a shared inbox. Some come from would-be distributors, some from existing distributors chasing stock, and some from end customers who should have gone to their local dealer. One operations coordinator sorts them by hand.
Last quarter a strong distributor application sat unread for nine days because it looked like a routine stock question. Priya wants enquiries sorted and routed faster. She does not want software replying to prospective partners in her company's name without anyone checking.
This is the kind of problem people bring to Make. The fit is plausible, but it depends on details we could not verify, starting with whether her form and inbox can connect at all.
What Make is, in plain terms
Think of a scenario as a recipe for moving information. Data comes from one tool, Make reshapes it, and passes it to another tool. How a scenario is started, scheduled or triggered was not described in the text we retrieved, so confirm that in Make's documentation.
Three layers are worth separating:
- Ordinary automation. Moving and reformatting data between apps. This is the documented foundation of scenarios (source).
- AI features. The pricing FAQ mentions Make AI Provider features that can cost more credits (source). The retrieved text did not say what those features do, which models they use, or how accurate they are.
- Your own judgment. Deciding what counts as a priority enquiry, who owns it and what a reply should say. No platform supplies this.
Many "AI automation" projects are mostly layer 1 with a small AI step in the middle. Knowing which part is which helps you predict cost and risk.
Documented features worth understanding
Scenarios that connect applications
A scenario links applications so data moves between them (source). The business value is fewer copy-and-paste handoffs. The retrieved evidence does not list supported apps, so do not assume your form tool, inbox or CRM works with Make until Make's own documentation confirms it.
Data transformation between steps
Scenarios can transform data before sending it on (source). Typical goals are splitting a name field, standardizing a country or trimming a message before it reaches a spreadsheet. Our source did not describe the specific transformation tools, so check which operations exist before designing around one.
Credit-based billing
Credits are Make's billing unit, and most actions consume one (source). For planning, this means steps per run and runs per month drive cost. Whether checks that find no new data also consume credits was not covered in our source. Ask, or test on a small scale, before estimating.
Make AI Provider features at a higher credit cost
Some advanced Make AI Provider features consume more than one credit (source). We could not confirm how many, which features, or whether you can connect your own AI provider account. Treat any AI step as a cost line to confirm, not a fixed assumption.
Who Make may suit
Make may suit an operator who:
- Already uses several online tools and spends real hours moving information between them.
- Can describe a process as clear steps with clear exceptions.
- Will own a scenario after it is built: watching failures, updating rules and checking credit use.
- Wants mostly rule-based handling with occasional AI help, and people making final calls.
The requirement is less technical skill than a process you can describe precisely.
Potential applications across business roles
These are proposed uses, not documented case studies. Each depends on your apps being supported.
- Sales and partnerships: Sort inbound partner or distributor enquiries into categories for the right owner, with a person approving any reply.
- Operations: Copy order or booking details into a tracking sheet in a consistent format, so a manager reviews exceptions instead of retyping.
- Customer service: Tag requests by type for a human queue. Sensitive cases stay with staff.
- Finance administration: Collect invoice details into one review list for a bookkeeper. Payments and approvals stay manual.
- Founder reporting: Gather weekly counts from several tools into one summary the founder reads and questions.
In each case the automation prepares information. A person decides.
A proposed workflow: distributor enquiry classification and routing
Illustrative only. It was not built or tested in Make. Steps marked "configuration unknown" depend on capabilities our evidence did not document.
Fictional input:
Name: Tomas R. | Company: Northgate Catering Supply | Region: North East | Message: "We supply 40 restaurants and want to stock your combi ovens. What are your distributor terms?"
Step 1. Capture. Bring each new enquiry from your form or inbox into a scenario. Configuration unknown: app support and how a scenario starts must be confirmed.
Step 2. Minimize data. Use transformation to keep only what routing needs: name, company, region, message and timestamp. Leave out phone numbers and attachments, so less personal data reaches later steps or any AI service.
Step 3. Rule-based pre-sort. Handle obvious cases without AI. For example, if your form asks "Are you a current distributor?", sort by that answer first. Configuration unknown: how Make expresses conditions was not in our source. This step saves credits and is easier to audit.
Step 4. Optional AI classification. For unclear messages, an AI step, if one fits your setup, suggests one category from a fixed list using the brief below.
Sample output (fictional):
Category: NEW_DISTRIBUTOR | Missing: none | Reason: Sender asks about distributor terms for a restaurant supply business.
Step 5. Handle missing data. If region is blank or the category is UNCLEAR, the record goes to a "needs review" list for the coordinator. Nothing is guessed or dropped.
Step 6. Route for human approval. The enquiry and suggested category go to the owner's tracking sheet or queue. Nothing is sent to the enquirer automatically. The owner confirms the category and writes or approves any reply.
Step 7. Pilot for four weeks. Each week, record:
- Share of AI suggestions the owner changed.
- Count of enquiries in "needs review".
- Days from arrival to first human reply, against your manual baseline.
- Credits used against your estimate.
Step 8. Stop route. Pause the AI step and return to manual sorting if owners change more suggestions than the threshold you set before starting, any enquiry goes missing, a sensitive message is misrouted, or credit use exceeds budget. Fix the cause before resuming.
Copyable classification brief
Role: You sort inbound business enquiries for a kitchen equipment supplier.
Task: Assign exactly one category from this list:
NEW_DISTRIBUTOR, EXISTING_DISTRIBUTOR, END_CUSTOMER, SUPPLIER_OR_VENDOR, UNCLEAR.
Rules:
- Use only the text provided. Do not infer facts that are not stated.
- If the message fits more than one category or none, return UNCLEAR.
- Do not draft replies or make commitments about pricing or terms.
- List any of these fields that are missing: company, region.
Output format:
Category: <one value>
Missing: <fields or none>
Reason: <one short sentence with no personal data>
The brief limits the AI to labelling. Replies and commercial terms stay with people.
Costs and limits
Numeric price not verified. The retrieved text did not include plan prices, and the English-language pricing URL was unavailable. Check the official pricing page directly. We could neither confirm nor rule out a free option.
Documented: usage is metered in credits, most actions take one, and some Make AI Provider features take more (source). Your cost will depend on how often scenarios run, steps per run and AI use.
Other limits:
- We could not confirm integrations, AI capabilities, data handling terms or user permissions.
- AI labels can be wrong. Plan for review.
- A scenario is a small system someone must maintain. Changes to your form or tools can break it.
Who should skip Make for now
- Businesses with a handful of enquiries a week. A shared inbox and a named owner may be enough.
- Teams that cannot name who will monitor failures.
- Anyone hoping to send AI-written replies to partners or customers without review.
- Organizations handling sensitive personal data that have not reviewed Make's data processing terms.
FAQ
Do I need to code to plan a Make scenario?
Planning needs process clarity more than code. Write the steps, categories, owners and exceptions on paper first. If you cannot describe the exception path, the scenario is not ready.
Should every enquiry go through the AI step?
No. Let clear cases follow fixed rules and keep AI for ambiguous text. Behavior stays predictable and the more expensive steps run less often.
How do I know the pilot worked?
Compare against the baseline you recorded before starting: fewer stalled enquiries, a correction rate you can live with, and spend within budget. Without a baseline you cannot judge the result.
Practical next action
This week, export your last 50 enquiries and sort them by hand into the five categories in the brief. Note how many were unclear and how long each waited for a reply. Then confirm in Make's official documentation that your form or inbox is supported and how AI steps are billed. Build only after both checks.
Methodology: AI-assisted draft based on official Make documentation retrieved on 2 November 2026. Two official pages were unavailable at that time. The story, workflow and sample data are illustrative, not a hands-on test. ShiftToAGI is independent and not affiliated with Make.