What Should a Nigerian Business Automate First? A Prioritisation Framework
The wrong first automation is expensive in two ways: the money spent, and the confidence lost when the team concludes that "automation does not work here". The right first automation is small, visible and quickly proven, and it buys permission for the second and third.
This article gives you a repeatable way to rank candidates rather than a single universal answer, because the correct first choice differs between a pharmacy in Ibadan, a logistics company in Lagos and a private school in Enugu. It ends with the typical first wins by business type and a worked hypothetical example.
Why the first automation matters more than the fifth
The first automation is a proof of concept for the whole company. It teaches the team what automation feels like, shows the owner what it costs and returns, and sets the pattern of writing rules, testing exceptions and measuring results. If it works, the second workflow is easier to approve and build. If it fails, the budget disappears and the topic is closed for a year.
That is why the choice should be driven by likelihood of success and visibility of results, not by the size of the pain. The most painful process in the company is often the hardest to automate, and it should usually be the third or fourth project, once the team has the habit.
The six-factor scoring framework
Score each candidate task from 1 (low) to 5 (high) on the six factors below. Factors one to four are benefits; factors five and six are penalties. Subtract the penalties from the benefits to rank candidates.
| Factor | What you are scoring | 5 means | 1 means |
|---|---|---|---|
| Frequency | How often the task runs | Many times a day | Monthly or less |
| Time consumed | Staff hours per week across everyone who touches it | 10+ hours weekly | Under an hour |
| Cost of errors | What a mistake costs in money, stock or customer trust | Frequent, costly errors | Rare, cheap errors |
| Rule clarity | Whether everyone would describe the process the same way | Fixed rules, no debate | Every case is different |
| Risk if it fails (penalty) | Damage if the automation misfires | Money lost or customers offended | Minor inconvenience |
| Dependencies (penalty) | How many other systems or people must change first | Needs new software and data cleanup | Works with what you have |
Score = Frequency + Time + Error cost + Rule clarity − Risk − Dependencies. A candidate scoring 12 or above is a strong first automation. Anything below 8 should wait.
A useful tie-breaker: prefer the candidate whose result customers or the owner will notice within a week. Automated receipts are noticed the same day; automated month-end reports are noticed in a month.
Common candidates ranked for Nigerian SMEs
The table below applies the framework to tasks that appear in almost every Nigerian SME audit. The scores are typical, not universal; run your own numbers.
| Candidate task | Freq | Time | Error cost | Rules | Risk | Deps | Score |
|---|---|---|---|---|---|---|---|
| Payment confirmation and receipt | 5 | 4 | 4 | 5 | 2 | 2 | 14 |
| Order or enquiry capture from WhatsApp into a record | 5 | 5 | 4 | 4 | 2 | 2 | 14 |
| Appointment and delivery reminders | 4 | 3 | 3 | 5 | 1 | 1 | 13 |
| Answering repeated questions (price, delivery areas, hours) | 5 | 4 | 2 | 4 | 2 | 2 | 11 |
| Invoice generation from a rate card or order | 4 | 4 | 3 | 4 | 2 | 2 | 11 |
| Low-stock alerts and reorder lists | 3 | 3 | 4 | 4 | 2 | 3 | 9 |
| Daily sales summary to the owner | 5 | 2 | 2 | 5 | 1 | 3 | 10 |
| Payment follow-up on overdue invoices | 3 | 3 | 4 | 3 | 3 | 2 | 8 |
| Payroll calculation | 1 | 3 | 5 | 3 | 5 | 3 | 4 |
| Credit approval for new customers | 2 | 2 | 5 | 1 | 5 | 3 | 2 |
Two observations from this table:
- The top candidates are all about money coming in and customers hearing back quickly. That is not a coincidence: those are the processes with clear rules and instantly visible results.
- Tasks with high error cost but low rule clarity, such as credit approval, look attractive because they hurt, but they are poor automation candidates until the rules are agreed.
What usually should not be automated first
Some tasks should be deliberately postponed, even when the owner is tired of them:
- Payroll. Infrequent, high-risk, and often entangled with informal arrangements that nobody wants to write down. Automate it once the rules are clean, and use established payroll software rather than custom scripts.
- Refunds and credit notes. A misfiring refund automation loses money directly. Keep a human approval step.
- Anything that depends on data you have not cleaned. If customer records are duplicated across three phones and a spreadsheet, an automation that reads them will produce duplicated messages. Clean first.
- Marketing broadcasts. Tempting because tools make them easy, but the WhatsApp Business Platform enforces opt-in and quality rules, and a badly targeted campaign damages the sender's standing. Start with transactional messages.
- Multi-department processes. Order-to-cash across sales, warehouse and finance is a great second-year project. As a first project it has too many dependencies.
Typical first automations by business type
Different Nigerian businesses tend to find their first win in different places. These are patterns, not prescriptions.
| Business type | Typical first automation | Why it scores well |
|---|---|---|
| Retailer or online seller | Order capture from WhatsApp or Instagram into an order record, with automatic payment confirmation | Highest frequency, clear rules, immediately visible |
| Professional services (law, accounting, consulting) | Enquiry intake form with automatic acknowledgement and internal assignment | Reduces lost leads; low risk |
| Clinic, salon or any appointment business | Booking confirmations and reminders | Cuts no-shows; simplest rules of all |
| School | Fee invoice generation and payment confirmation per pupil | Seasonal but high value; rules are fixed |
| Restaurant | Order confirmation and kitchen notification for delivery orders | Frequency and speed matter most |
| Logistics or delivery | Shipment status updates to customers on each scan or handover | Cuts "where is my package" calls |
| Manufacturer or distributor | Payment confirmation and dispatch trigger | Links cash to stock movement |
| Real estate | Enquiry capture and viewing appointment reminders | Slow response loses buyers |
Industry-specific detail is covered in the dedicated articles on automation ideas for Nigerian retailers, restaurants, hotels, schools, logistics companies and others in this cluster.
What changes in Nigeria
Several local realities shift the scoring compared with a generic framework.
Bank transfer is the default, and confirmation is manual by tradition. This is why payment confirmation sits at the top of most Nigerian lists. Virtual account numbers and gateway webhooks from providers such as Paystack, Flutterwave or Monnify turn a nightly bank-app ritual into an instant trigger, with an exception path for mismatched transfers.
WhatsApp is the front door. Enquiries, orders and complaints arrive on WhatsApp regardless of what else you build. Capturing that traffic into structured records scores highly because it is frequent and because the alternative (staff memory) is the most common source of errors.
Connectivity affects the risk score. A task whose automation must complete in real time on a mobile connection carries a higher risk penalty than one that can queue and sync. Reminders can wait ten minutes; a POS sale cannot.
Currency affects the dependency score. If the automation requires a new USD-priced platform, treat that as a dependency penalty, both for the cost and for the risk of the subscription lapsing when a card payment fails.
Owner-centred approvals inflate the "time" score for the wrong tasks. The MD spending an hour a day approving small purchases looks like a candidate for automation. Usually the fix is a threshold rule, not software.
Example (hypothetical): scoring eight candidates at an Abuja furniture company
Example (hypothetical): A made-to-order furniture company in Abuja with 14 staff sells through Instagram and WhatsApp, takes 50–70% deposits by bank transfer, and delivers within the FCT and to Lagos by third-party haulage. The owner lists eight candidate tasks from a two-week audit and scores them with the team.
| Candidate | Score | Decision |
|---|---|---|
| Deposit confirmation and receipt to customer | 14 | First |
| Enquiry capture from Instagram DMs and WhatsApp into a lead sheet | 13 | Second |
| Production status updates to customers ("frame complete", "upholstery done") | 12 | Third |
| Delivery scheduling and haulage booking | 8 | Later; depends on haulier's availability |
| Quotation generation from a price list | 10 | Fourth, once the price list is standardised |
| Supplier reorder for foam and fabric | 7 | Later; stock data not reliable yet |
| Balance-payment follow-up before delivery | 9 | Fifth; pair with production updates |
| Custom design approval with the client | 3 | Never; judgement and relationship work |
The first automation is deposit confirmation: a virtual account per order, so an incoming transfer marks the order as confirmed, sends a receipt on WhatsApp and creates a production job. It took the team two weeks with a workflow platform and existing invoicing software. The second and third follow the same pattern and reuse the order record. The example is illustrative, not a client result.
Indicative cost of a first automation
Because the first automation should be small and low-dependency, it is also usually cheap. Indicative 2026 ranges; actual quotes vary with scope, vendor and exchange rate.
| Route | One-off | Recurring |
|---|---|---|
| Built-in features of tools you already pay for | Staff time | Existing subscription |
| Workflow platform plus existing tools, done in-house | ₦0–₦100,000 (staff time, training) | USD subscription per month |
| Workflow platform set up by an agency or freelancer | ₦100,000–₦1,000,000 | USD subscription per month |
| WhatsApp Business Platform onboarding with a simple flow | ₦150,000–₦1,500,000 | Per-conversation charges in USD |
| Small custom integration (for example, payment webhook to your own system) | ₦300,000–₦2,000,000 | Hosting and maintenance |
The purpose of the first project is learning, so a budget that keeps it under ₦1,000,000 all-in is reasonable for most SMEs. Compare two or three written quotations on the same one-page specification, and get the full picture of costs from the article on how automation can reduce business costs in Nigeria before deciding what to spend.
How to move from the first automation to a roadmap
Once the first workflow is live and measured, the same scoring sheet becomes a roadmap:
- Re-score the remaining candidates. Dependencies often fall because the first automation created a clean record (the order, the customer) that later automations can reuse.
- Group candidates that share a data source. Payment confirmation, receipts, balance reminders and dispatch triggers all hang off the order record, so they are cheaper together.
- Schedule one workflow per four to eight weeks. Faster than that and the team cannot absorb the change.
- Revisit the "not yet" list every quarter. Payroll and credit rules may be clean enough by then.
- When three or more workflows depend on each other, step up from task automation to process design; the article on business process automation in Nigeria explains that method.
Mistakes in choosing what to automate first
- Choosing the most painful task instead of the most automatable one. Pain is a signal to investigate, not a ranking. Payroll pain usually means unclear rules, not missing software.
- Ignoring the risk penalty. A first automation that can send the wrong amount, the wrong message or the wrong refund is not worth the saved hours.
- Ignoring dependencies. If the candidate needs a CRM you do not have and data you have not cleaned, it is a third project pretending to be a first.
- Automating a task nobody outside the office will notice. Internal report generation may score well, but it will not build support for the programme the way faster receipts and confirmations do.
- Letting the tool choose the task. Signing up for a platform and then looking for something to automate leads to trivial automations and a monthly USD bill.
- Scoring alone. The owner's view of frequency and time is often wrong. Score with the staff who do the work.
Conclusion
The first thing a Nigerian business should automate is the task that is frequent, rule-based, low-risk and low-dependency, which in practice usually means payment confirmation, WhatsApp order or enquiry capture, and customer reminders. Score candidates on frequency, time, error cost, rule clarity, risk and dependencies, choose the highest scorer that customers or the owner will notice quickly, and postpone payroll, refunds and credit decisions until the rules are clean. A small, successful first project buys permission and creates the data foundation for everything after it.
If you have a shortlist and want help scoring it or building the first workflow, Linestech advises Nigerian businesses on automation priorities and implements the workflows, integrations and custom tools that follow.
Frequently asked questions
Is customer service a good first thing to automate?
Partly. Automatic acknowledgements, business-hours messages and answers to a handful of fixed questions (price list, delivery areas, opening hours) are excellent first automations because they are frequent and low risk. Full customer-service automation with routing, ticketing or AI answers is a later project that depends on clean product data and a defined escalation route to a person.
Should a very small business with two staff automate anything?
Yes, but at the lightest tier. Automated receipts and payment confirmation through a payment gateway, WhatsApp Business App quick replies and a booking or invoicing tool's built-in reminders cost little and remove the tasks that keep a two-person business working late. Custom software rarely makes sense at that size.
How do I score a task if my team disagrees on the rules?
That disagreement is itself the answer: rule clarity scores 1 or 2, which pushes the task down the list. Agree the rule first, run it manually for a month, and re-score. Automation cannot resolve a policy debate; it can only enforce whichever version gets built.
Can AI change what I should automate first?
AI widens the set of candidates by handling unstructured inputs, such as reading a mixed English-and-Pidgin WhatsApp message and extracting an order, or classifying enquiries. It does not change the framework: those tasks still need clear downstream rules, a review step and a defined exception route. For most SMEs, AI-assisted capture becomes the second or third automation, once the structured record it feeds exists.
How many tasks should I automate in the first year?
Three to six well-chosen workflows is a realistic first-year target for an SME, one every six to eight weeks. The constraint is the team's ability to absorb change and keep each automation healthy, not the technology. A company with six working, measured, owned automations is in a far better position than one with fifteen half-working ones.
What if my first automation fails?
Diagnose before abandoning. The usual causes are unclear rules, an untested exception (wrong-reference transfers, offline periods), staff bypassing the system, or a personal account being used. Most first failures are fixable in days. If the task itself turns out to be unsuitable, drop down the scoring list and try the next candidate with the lessons applied.
Sources and further reading
Figures, platform rules and regulations change. These are the primary references behind this article and the places to check before you act on it.
