How to send WhatsApp messages from AutoCount (the no-wait way)

The Meta Embedded Signup wait is the same for every platform — two to four weeks, sometimes longer, depending on the review queue. The difference is whether you spend the wait filling in Meta's verification forms yourself, or whether the platform you are using already holds the Meta Business verification, in which case the queue is skipped and the business just has to prove the number is theirs. Here is the no-wait path for AutoCount.

The slow path: BSP + AutoCount + the wait

The standard advice is to pick a Business Solution Provider, sign a contract, get a number assigned or ported, and let the BSP handle the Meta verification. Most BSPs quote setup fees in the RM2,000 to RM5,000 range, plus a monthly platform fee, plus the Meta conversation fees, plus a per-message markup on top of Meta's. The total is not unreasonable. The timeline is. Six to eight weeks, conservatively, before the first message lands.

Then there is the integration. The BSP's API is a generic WhatsApp API — it does not know what your AutoCount sales ledger says about the customer on the other end of the message. The integration to AutoCount is yours: a developer, a project, a maintenance commitment. Most Malaysian SMEs end up with a half-built integration that sends templated messages and nothing more.

What the no-wait path actually is

minidesk holds the Meta Business verification. When you sign up and pick AutoCount as your connector, the Embedded Signup window opens, you pick the WhatsApp number you want to use, Meta verifies the number belongs to you (this part is automated and takes minutes), and the API credentials land in your minidesk account. The whole flow is twenty minutes, end to end. The Meta review queue is skipped because the queue is for businesses that have not been verified yet — your verification rides on minidesk's.

The second part of the no-wait path is that the WhatsApp message already knows what the AutoCount database says. When the AI sends a message to a customer, it reads the AutoCount sales ledger, finds the customer's open invoices, picks the right one for the conversation, and writes a message that includes the invoice number, the amount, and the due date. When the customer replies, the AI reads the reply, updates the AutoCount follow-up notes, and tells your bookkeeper the thread is live. Your bookkeeper does not log into AutoCount. She does not log into WhatsApp. She reads the AI's summary at 5pm and decides which threads need a human reply.

What the conversation looks like in practice

It is 11:42am on a Wednesday. A customer in George Town — Encik Tan, who runs a hardware retail shop — sends a WhatsApp: "Saya nak check balance." The AI reads the phone number, identifies Encik Tan from the AutoCount debtor master, pulls the open balance, and replies in Bahasa Malaysia: "Selamat tengahari Encik Tan, baki terbuka anda adalah RM3,847.50 merentasi 2 invois — INV-2025-06234 (RM1,420, due 18 July) dan INV-2025-06891 (RM2,427.50, due 2 August). Sila beritahu saya jika perlukan penyata." The reply lands at 11:42:08am. Encik Tan reads it on his lunch break, replies with a payment confirmation, and the AI marks the follow-up note in AutoCount. By the time Encik Tan's payment lands the next morning, the bookkeeper has already seen the thread and approved it.

“The wait was the part I could not afford. I was losing two sales a week to customers who messaged in the afternoon and got no reply until the next morning. Now the reply is the same minute, and it already knows the order.”

— Linda, who runs a chain of three retail shops in KL

What to ask the vendor if you are shopping for a real ERP

If you have decided the word "ERP" applies to your business, the vendor conversation is the part that determines whether the next two years are a success or a write-off. Six questions worth asking, in the order the answers matter. (1) What is the implementation timeline, in weeks, from signed contract to first user login? (2) What is the all-in cost over 3 years, including licence, implementation, training, support, and the integrations you will need? (3) How does the system handle multi-company consolidation across the entities you operate today, and the entities you might operate in 3 years? (4) What is the cost of adding a 50th user, a 100th user, a 500th user? (5) How is the data isolation enforced — at the database level, in the application code, or in a policy document? (6) What does the support queue look like at 11pm on a Saturday, and is it staffed by humans or by an AI that does not know your business?

The answer to question 5 is the one that matters most for the long-term safety of your data. An ERP that enforces the data isolation at the database level is an ERP that cannot accidentally leak one company's data to another. An ERP that enforces it in application code is an ERP that will leak, eventually, when the application code changes. An ERP that enforces it in a policy document is an ERP that has not enforced it at all. The honest vendors will tell you which one they do, in one sentence, with a diagram. The dishonest ones will tell you they do "role-based access control," which is a different thing, and which is a polite way of saying the wall is in the prompt.

The honest case for staying on what you have

For the 80% of the businesses we talk to, the honest answer to "is it time for an ERP" is no. The honest answer is: the software you have is doing the job the software was bought to do, and the next ring of investment is not a bigger software, it is an assistant that makes the software you have do more. The bookkeeper does not need a new chart of accounts. She needs the chart of accounts she has to answer the WhatsApp that is sitting unread. The owner does not need a new dashboard. He needs the dashboard he has to read itself back to him in a sentence. The migration to a real ERP is a 6- to 18-month project that costs more than the software itself. The AI on top of what you have is a 20-minute install that pays for itself the first month. If you are the 80%, the AI is the right next step. If you are the 20%, we will tell you so — and we will tell you which ERP we would shortlist, because we have read the same industry analyses you have.

What a normal Tuesday looks like with the AI on

It is 7:42am on a Tuesday. The owner is at the counter, the staff are opening the shop, the WhatsApp is already pinging with overnight messages from customers. Before the AI, the morning was the same every day: open the ledger, check the bank balance, open WhatsApp, start answering. After the AI, the morning is one prompt and a coffee.

The owner opens the minidesk chat on the laptop. The AI reads the ledger, the bank feeds, the open WhatsApp threads, and replies in 4 seconds with a three-point summary: who paid overnight, who is past due, which customer follow-ups are queued for approval. The owner taps approve on the queued messages, the WhatsApp is up to date, the ledger is clean, the bank feed is reconciled. The morning is done by 7:46am.

That is the shape of a normal day. The bookkeeper opens her ledger dashboard once, around 10am, to do the human-shaped work — the reconciliation that needs judgement, the journal that needs a note, the month-end that needs a sign-off. The AI handles the rest: the WhatsApp, the reminders, the daily summary, the "any update?" reply. The owner handles the strategy. The bookkeeper handles the judgement. The AI handles the repetition. Everyone's day gets a little longer, and the customer gets a faster answer.

What it costs, in real money and in real time

The AI plan is a flat monthly fee — the number is on the pricing page, the quote is the quote, no per-message markup, no per-prompt surcharge, no enterprise tier. The translator for desktop ledgers is included. The OAuth for cloud ledgers is included. The WhatsApp Business API access is included. The Meta conversation fees are the same fees every platform pays, and we do not add a per-message markup on top. The only other line item is the AI usage, which is metered in the same units every modern AI tool uses, and which your bookkeeper will not notice unless she is doing something unusual.

The time saving is harder to put a number on, because the time is not a single block — it is a thousand two-minute savings across a month. The Monday morning phone tag that used to take three hours. The Friday afternoon invoice chase that used to take ninety minutes. The "any update?" reply that used to take fifteen minutes per customer. The end-of-day reconciliation that used to take an hour. The end-of-month report that used to take a day. The AI takes each one of those from minutes to seconds, and the sum is the difference between a business that is drowning in WhatsApp and a business that is using WhatsApp to grow.

What the install looks like, in 5 steps

The whole setup is twenty minutes from signup to the first message. Five steps, all from the same browser window:

1. 1Sign up at minidesk.co and pick AutoCount as your connector. The on-prem and the AutoCount Cloud editions both work.
2. 2Install the translator on the PC that runs AutoCount. The translator speaks Firebird and SQL Server, and reads the same database the AutoCount desktop client reads.
3. 3Connect your business WhatsApp number through Meta Embedded Signup. Meta verifies the number is yours, the API credentials land in your account, the window opens.
4. 4Authorise the AI to read the AutoCount database. The translator enforces the same row-level permissions your AutoCount user has — the AI cannot see a customer your bookkeeper cannot see.
5. 5Ask the AI to send the first message. The first message is usually a test to your own number, so you can see the layout, the language, and the speed. The second message is usually a real one.

If your business is already on AutoCount, the WhatsApp path is the part of the AI you will use the most, and it is the part that takes the least time to set up. Twenty minutes from signup to the first message. The bookkeeper does not have to do anything different. The customer does not have to download an app. The message lands in the WhatsApp they already have open.

→ See the AutoCount integration: /integrations/autocount