For software alone, connecting two business apps can cost $0 extra with an included native integration, from about $9 a month with Make, or $29.99 monthly with Zapier Professional. Custom API work needs a scoped quote. Add setup, monitoring and maintenance to compare the four options fairly.
The useful price is the cost of moving your actual records correctly. Sending a new enquiry one way is a different job from keeping bookings, cancellations and payments consistent in both systems. An integration that misses cancellations can create more work than copying the information yourself.
Describe the transfer before asking for a price
An API is a way for one piece of software to request information or actions from another. An integration uses that access, or another supported connection, to move information between apps. You do not need to understand the code. You do need to describe what should happen when the information changes.
Write one sentence with a trigger, an action and a boundary. For an illustrative wedding planner: “When an enquiry becomes a confirmed booking, create one planning record with the booking reference, event date and package; do not copy private consultation notes.” A trigger starts the transfer. An action is something it does.
Add the awkward cases before shopping. What happens if the booking reference is missing? If a staff member changes the date twice? If an enquiry is entered twice? If the receiving app is unavailable? A supplier who prices only the successful first transfer has not priced your whole requirement.
Spend an initial 45 minutes listing the fields, then inspect ten recent records together. This is a suggested planning allowance, not a standard implementation time. If colleagues disagree about which app holds the correct event date, resolve that first. The process-mapping tutorial helps turn a vague request into a sequence someone can price.
Keep a manual baseline. An illustrative barber shop copies 30 membership changes a month, taking two minutes each. That is one hour. At an assumed internal time value of $24 an hour, the present workload is $24 a month. A $30 subscription cannot pay for itself through this task alone, even before setup and checking.
Option one: use the connection inside either app
A native integration is a connection offered within one of the two products. Check both vendors' integration lists and the documentation for your exact plans. Do not infer compatibility from a logo. A booking app might offer contact creation without offering booking changes, cancellation updates or custom fields.
Ask three price questions: is the connection included, does either app need a higher plan, and does usage attract a separate fee? The answer depends on the two vendors. There is no honest universal “native integration price”. Put the actual incremental charges in your comparison, including any seats that must move to a higher plan.
For an illustrative nail salon, suppose a vendor quote puts the necessary plan upgrade at $12 a month. This is a made-up quotation for the calculation, not a product price. If it applies to four staff accounts, the incremental bill is $48, not $12. If the upgrade is account-wide, the calculation changes again.
The setup can still require decisions. Map “client reference” to “customer ID”, decide whether blank values overwrite existing ones, and check whether records created before activation will transfer. Use dummy bookings until you know which actions send notifications. A test booking that reaches a real client is a real booking problem.
This route deserves the first look when it covers the exact transfer and someone can inspect failures. It becomes less attractive when the team must maintain a second workaround for every changed booking. Compare the total work with native integrations versus Zapier, rather than assuming built-in means complete.
Option two: price Zapier by actions, not enquiries
Zapier's Professional plan starts at $29.99 a month on monthly billing, or $19.99 a month billed annually, for 750 tasks. These are list prices in USD. A task is a successful action step. Triggers and filter steps do not count as tasks, so the number of incoming enquiries is not automatically the billable total.
An illustrative personal trainer receives 180 package enquiries a month. A proposed workflow creates a contact, adds a follow-up task and records the enquiry in another app. If those are three successful billable actions, the expected usage is 180 × 3 = 540 tasks. A busy month with 280 enquiries becomes 840 tasks before any other workflows.
That difference matters. Do not promise that the 750-task allowance will cover the busy month, or guess what the next allowance costs. Check the current pricing for the required volume and what happens at the limit. Leave room for testing, backlogs and extra successful actions caused by your recovery process.
Zapier's Free plan has 100 tasks a month and two-step Zaps. A Zap is its name for an automated workflow. That can be useful for a small, simple trial, but it is not the budget for the trainer's three-action design. Free-plan polling checks every 15 minutes; Professional checks every two minutes. Polling means checking an app repeatedly for new information.
Those intervals are not a guarantee that every end-to-end transfer finishes within that time. The source app, connection method and destination can also affect delay. Define acceptable waiting time from the customer's point of view. An enquiry copied within a few minutes may be fine; a promise to reserve the last appointment needs stronger coordination.
Use the official Zapier task-counting guidance when checking a proposed design. Ask the implementer to show the task estimate for one normal record, one changed record and one failed record that is later processed successfully.
Option three: cost Make from the actual run
Make has a free allowance of up to 1,000 credits a month. Its entry paid option starts from about $9 a month for 5,000 credits on monthly billing. Make restructured its plans in 2026, so use the current credits-based pricing page rather than relying on older plan names or screenshots.
Credits measure usage, while operations describe the work performed by modules in a workflow. A module is a processing step. Do not assume that a Make credit equals a Zapier task. Count the modules, the records passing through them and any different charging rules for the particular modules you choose.
Consider an illustrative yoga studio transferring registrations. A trial of 20 representative registrations reports 120 credits in the usage history. That is six credits per registration for this measured route. At 600 registrations, the same pattern suggests 3,600 credits. Add a separately measured allowance for checks, retries and other scheduled work.
The studio should test a registration with three attendees as well as one with a single attendee. A workflow might process each attendee separately. A price based only on single-attendee tests can be misleading even when the monthly number of bookings remains steady. Multiply by the records actually processed, not the number of forms submitted.
Make routers use no credits, and a bundle stopped by a filter uses none. A bundle is a packet of data moving through the workflow. That does not erase usage already incurred earlier in the run. Place checks thoughtfully and read the execution history before concluding that rejected records are free overall.
Make is worth testing when its available modules cover your steps and the person maintaining it can follow the flow. Its entry price does not establish that it will cost less for your workload. Give Zapier and Make the same transfer requirements, sample records and error cases before comparing their bills.
Option four: commission a custom API connection
A custom connection is code written for your particular transfer. Its main cost is usually the work you specify: mapping records, handling access, checking failures, testing changes and documenting ownership. Avoid asking for a universal price per API. Two apps with clear, supported interfaces can still have complicated business rules between them.
Ask whether both vendors permit and support the required API access on your plans. Find out what usage limits apply and whether a test environment is available. A rate limit is the number of requests a service allows within a period. The supplier should explain how a backlog is handled when requests must wait.
Here is an illustrative budget calculation, not a market rate or quotation. Assume 24 implementation hours at $80 an hour: $1,920 one-off. Add a hypothetical $15 monthly hosting allowance and two maintenance hours at the same rate: $175 a month. In the first year, that is $1,920 + 12 × $175 = $4,020.
Every assumption needs replacing with a quotation. Hosting might be included elsewhere. Maintenance might be charged only when needed. A connection involving several record types, permission checks and two-way updates could require much more work. The point of the example is to expose the cost categories, not suggest a normal industry price.
A tattoo studio illustrates the reason to consider custom work. It wants booking references and appointment status to transfer while consent forms and private notes stay out. If available integrations cannot enforce that boundary, a custom design may be worth investigating. First ask the vendors whether a supported configuration already solves it.
Make delivery concrete. Require a record of the mapped fields, the accounts used, failure notifications, how to pause the transfer and how to recover a missed booking. The business should own the relevant accounts and receive enough documentation for another competent supplier to maintain the connection.
A wedding planner compares the same job four ways
The main worked example is an illustrative three-person wedding planning business moving confirmed bookings from an enquiry app into a planning app. It handles 120 new bookings or relevant changes a month. Each manual transfer currently takes four minutes, including checking the destination. That is eight hours monthly, valued internally at $30 an hour: $240.
The business needs a one-way transfer, a stable booking reference and a visible list of exceptions. It does not need AI to interpret the records because the source already supplies clean fields. Adding an AI step would introduce another cost and another possible error without solving an identified problem.
The table uses hypothetical implementation estimates to demonstrate a comparison. Only the stated Zapier and Make subscription figures are product list prices. Internal time is valued at $30 an hour throughout. The custom supplier estimate uses $80 an hour. Base subscriptions for the two business apps are excluded because they stay the same.
| Route | One-off planning allowance | Monthly incremental allowance | First-year total |
|---|---|---|---|
| Included native connection | 4 internal hours: $120 | No extra software; 1 checking hour: $30 | $480 |
| Zapier Professional | 8 internal hours: $240 | $29.99 software; 1 checking hour: $30 | $959.88 |
| Make entry paid option | 10 internal hours: $300 | About $9 software; 1.5 checking hours: $45 | About $948 |
| Custom connection | 24 supplier hours plus 4 internal hours: $2,040 | $15 hosting, $160 maintenance, $30 checking: $205 | $4,500 |
The native row assumes the connection is genuinely included and covers all required fields. The Zapier row assumes three billable actions per event, giving 360 tasks a month. The Make row assumes a measured four-credit route, giving 480 event-related credits, plus the trigger's own checks explained below. These are designs to verify, not claims about unnamed apps.
Why pay for Make when 480 looks well below the free 1,000 credits? Because the trigger costs credits too. Make charges a credit every time a scheduled trigger checks for new records, even when nothing has arrived. Checking every 15 minutes is 96 checks a day, or about 2,880 credits in a 30-day month, which exceeds the free allowance on its own. Add the 480 event credits and the route needs about 3,360, inside the entry plan's 5,000.
The schedule is therefore a pricing decision. An hourly check uses about 720 credits a month, so 720 plus 480 would still exceed the free tier, but only just. A booking that waits up to an hour may be acceptable for a planning record and unacceptable for a same-day enquiry. Zapier works differently: its triggers never use tasks, however often they poll. Keep the free Make option on the shortlist only if a schedule you can live with keeps total usage under 1,000.
The manual baseline is $2,880 of internal time per year. Subtracting the totals gives possible first-year capacity value of $2,400 for native, $1,920.12 for Zapier and about $1,932 for Make. The custom route costs $1,620 more than the baseline. These figures assume the proposed checking times prove accurate and the previous transfer work disappears.
This is not cash profit. Staff salaries may stay unchanged, and the recovered hours only help if the business uses them. Nor does the cheapest row automatically win. If native cannot transfer cancellations, it fails a requirement and should be removed from the shortlist rather than awarded a low price for an incomplete job.
The close Zapier and Make totals show why software price alone can mislead. In this illustration, Make's lower subscription is almost offset by greater internal setup and checking allowances. Change those measured times and the result changes. Keep a small pilot before committing to a year of subscription or support.
Recalculate when booking volume drops
Suppose the planner enters a quiet period with only 30 events a month. Manual copying falls to two hours, worth $60 at the same internal rate. If the Zapier route still needs one checking hour and the $29.99 subscription, its monthly cost is $59.99 before recovering any setup cost. The original eight-hour saving no longer describes this month.
This does not mean the connection must be switched off each quiet month. Reliability, continuity and the coming busy period may justify keeping it. It does mean the annual decision needs a realistic mix of busy and quiet months. Use the last twelve months of event counts if you have them, and label any forecast separately.
Apply the same discipline to growth. Doubling volume does not necessarily double maintenance, but it can cross a usage allowance or expose a new exception pattern. Ask for a second estimate at twice normal volume. Keep the record size and number of actions explicit, so a supplier can explain which cost changes and which stays fixed.
Five details that can change a quotation
Historical records are a separate piece of work
An illustrative yoga studio has 2,400 old member records. Its request says “connect new memberships”, but the owner also expects the old records to appear. Price the initial transfer separately. Agree whether inactive members are included, which duplicate wins and how many records should exist afterwards. A one-off backlog can exceed a monthly usage allowance.
Two-way changes need an owner for each field
A personal trainer changes an appointment in the calendar while an assistant edits it in the client app. Which wins? “Latest change” sounds easy until delayed transfers arrive out of order. Prefer one source for the appointment date and a separate source for payment status. Write that division in the brief before accepting a two-way sync quote.
Retrying must not create a second booking
An illustrative barber shop sees a timeout after the receiving app has already created a membership. The workflow retries and creates another. The customer now appears twice. Require the transfer to check a stable source reference before creating again. Ask the supplier to demonstrate this with a test interruption, including the record of what happened.
Access changes can stop an otherwise sound transfer
A nail salon's connection was authorised by a manager who later leaves. The handover needs an account owner, a way to renew access and a named person who receives failures. Check the vendor's supported account arrangements rather than sharing a staff password. Include access review in the monthly work allowance.
AI extraction changes both cost and checking
A wedding planner may receive free-text requests such as “same package as last time, but move the ceremony later”. That is not a clean date field. An AI step might draft an interpretation, but a person must resolve ambiguity. Zapier's AI steps can consume different task amounts depending on the chosen tier, so do not apply ordinary action arithmetic blindly.
Keep these cases distinct in the quote. A structured field transfer, an AI interpretation step and permission to make a customer commitment are three different responsibilities. If the information is unclear, the correct output may be an exception for staff, not an automatically updated booking.
Use a quotation brief that makes exclusions visible
Send each supplier the same short brief. Include normal volume, peak volume, acceptable delay and example failures. Otherwise one price may include monitoring and another only the initial setup. The no-code versus custom-build comparison can help when a supplier proposes a larger build than the transfer appears to need.
Job: confirmed booking to planning record
Direction: enquiry app to planning app only
Match key: booking_reference
Fields: reference, date, package, status
Exclude: private notes, consent forms, payment card details
Volume: 120 events monthly; test at 240
Duplicates: update the matching record; do not create another
Failure: show the reference in an exception list
Acceptance: counts reconcile; cancellations transfer; no duplicates
Quote separately: setup, app upgrades, usage, support, handover
Explain: pause, recovery, ownership and future changesA useful response might read: “Eight setup hours include field mapping, twenty test records and one handover. Historic records and two-way updates are excluded. One hour of monthly review is an internal responsibility.” That is illustrative wording. It tells the owner what to compare and what still needs budgeting.
A less useful response says “full integration included” with no field list or support boundary. Ask for the detail before treating it as a fixed price. Clarify who investigates when one vendor blames the other, whether support covers business-rule changes and how extra work is approved.
Check the transfer and the bill during a small pilot
Allow a pilot to cover at least one full cycle of the work. For a weekly membership update, that means observing a real update and its corrections, not just creating one dummy member. A fortnight may be a reasonable planning window, but a monthly process needs longer before its routine operation is proven.
- Run controlled records. Include a normal booking, a changed date, a cancellation, a duplicate and a missing required field. Keep customer messages switched off in the test configuration where supported.
- Reconcile both ends. Count eligible source records and correctly matched destination records. A workflow marked successful is not proof that every field contains the right value.
- Inspect exceptions. Check that someone can identify the failed reference, understand the reason and recover it without repeating completed actions.
- Read actual usage. Compare the billable tasks or credits with the estimate. Record why a multi-attendee booking or retry used more.
- Time the checking. Include opening logs, correcting mistakes and answering staff questions. Replace the budget allowances with observed minutes.
If the transfer succeeds but the bill grows faster than volume, inspect loops, repeated searches and unnecessary updates. If the bill stays low but staff still copy records, the workflow may be missing an important branch. If no failures appear, deliberately create a safe test failure to establish whether alerts work.
Finish with a short acceptance note: which route was selected, monthly usage at current volume, who owns it and when costs will be reviewed. Keep the original manual instructions available for an outage. A connection is worth buying when its full operating cost fits the work it reliably removes, and the business knows how to keep it running.
Further reads
- When Does Zapier Get Too Expensive? Finding the Tipping Point — Find the usage patterns that push up automation bills.
- Zapier vs Make vs n8n for AI Automation: Which Fits Your Business? — Compare platforms against your maintenance skills.
- How to Set Spending Limits and Alerts on Pay-As-You-Go AI Tools — Set a budget before usage starts to grow.
- Business Data Backup Checklist Before You Connect AI Tools — Prepare recovery before connecting business records.
- Is It Worth Automating a Task You Only Do Once a Week? — The payback sum for weekly tasks, the factors beyond time that change the answer, and three worked cases: one yes, one partly, one no.
- Can AI Work With the Tools Your Business Already Uses? — The four ways AI connects to the software you run on, a 30-minute stack audit, an insurance broker's tools mapped, and workarounds for old systems.
- What Is an API? Why It Matters When You Buy Software — A plain-English explanation of APIs for people buying software, with the vendor questions and pricing traps that decide whether tools can connect.
- What Is a Webhook? Why Some Automations Run Instantly — Why some automations fire in seconds and others lag by minutes, how to tell which you have, and how to keep webhooks from failing silently.
- How to Stop Retyping Data Between Apps With AI Automation — Find every place your team copies data by hand, match each to the cheapest fix, and use AI extraction only where the source is an email, PDF or free text.
- Power Automate for Small Businesses: When It Beats Zapier — A decision guide for Microsoft 365 businesses: what your licence covers, the premium-connector line, how licences really work and three costed scenarios.
- How to Connect ChatGPT or Claude to Your Business Apps — ChatGPT apps, Claude connectors, custom MCP and Zapier routes compared, with a wedding planner's setup and the permissions to set first.
- iPaaS, Middleware or Custom Code: How Should Your Apps Connect? — iPaaS, middleware and custom code explained in plain English, with one real-shaped integration costed three ways over three years.
- How Much Does It Cost to Automate One Workflow With AI? — Every cost line in one AI-automated workflow, with a returns-email example costed on Zapier and Make, three budget scenarios and where bills creep up.
- What Does It Cost to Integrate AI Into Your Existing Software? — Four ways to add AI to software you already run, what each costs to build and run, and a letting agency's three quotes compared per request.
- AI Tools and AI Development: The Complete 2026 Guide — the AI hub, including every tutorial in the AI-for-business series.
Sources: Zapier pricing page and help article, How is task usage measured in Zapier?; Make pricing page and help pages on credits, operations and scheduling (checked September 2026). Labour, hosting and native-upgrade amounts are explicitly illustrative assumptions.