The five stages of a small business and what its back office has to withstand

It is the 3rd of the month and your accountant writes to you: eleven supplier invoices are still missing before the books can be closed. You remember paying at least eight. The XML files are scattered across three email inboxes.

At the bank there is a deposit of 46,400 pesos that nobody can match to a customer. And the only person who knows who owes you and how much is you. (Example scenario: it does not describe a real company.)

None of this is serious on its own. Taken together, it says something: your company has outgrown the way you run it.

One clarification before we start: Hábil did not work with Grupo Rotoplas and did not take part in any of the events told here. We tell them using what the company itself reported to the Mexican Stock Exchange and serious press, each fact with its reference number. When something is our reading, we say so.

In 1 minute

• The risk: running a 40-person company with the tools that worked when there were 5. According to Churchill and Lewis's five-stage model (1983), many founders who reach success fail when they grow [1].
• What inaction costs: in Mexico, 72.0% of the companies that did not get the largest loan they applied for were affected; above all, their expansion or the purchase of machinery was delayed (45.0%) [7].
• The case: Rotoplas, already with 500 employees in 1995, first put its processes (2003) and its measurement (2005) in order; then it rolled out one system in all its countries (2008) and on top of it centralized finance and collections (2009) [11].
• What you learn, in our reading: put your processes in order before buying the system. A common system is what later lets you centralize the back office of several sites or legal entities.
• What you have to decide on Monday: which stage you are in, whether you want to grow or stay as you are, and what to fix before the next one. At the end there is a list for your meeting.

In this article

1. What does a small business in Mexico have to work with today?
2. What does the Churchill and Lewis model say, in short?
3. Which stage is your company in?
4. Does your company stop when you are not there?
5. Why do you sell and still run short of money?
6. Grow or stay as you are?
7. Why does cash get tighter the more you sell?
8. How do you consolidate what you have gained without becoming slow?
9. What does the sequence look like in a Mexican company?
10. What changes in Mexico with the CFDI?
11. What breaks in your back office at each stage?
12. Which mistakes cost the most?
13. What should you take to Monday's meeting?
14. Key takeaways
15. What does Hábil do?
16. Frequently asked questions

What does a small business in Mexico have to work with today?

Most Mexican SMEs operate without financing, and many without planning their finances: an orderly back office is their cushion.

Two surveys by INEGI, Mexico's national statistics institute, show it. The first, ENAPROCE 2018, covers micro, small and medium-sized companies [5]:

• Only 8 out of 100 had financing in 2017 [5][6]: 7.6% of micro companies and 23.2% of small and medium-sized ones [5]. And 7 out of 10 would not accept a bank loan even if it were offered to them; of those, 57.9% said loans are expensive [5].
• Only 15.3% trained their staff in 2017 [5][6]. Among small and medium-sized companies, 55.1% [5].

What it means for you, in our reading: if you grow on your own cash, the money lost to forgotten collections or to unidentified deposits does not reach the expansion. And if almost nobody trains, your back office lives in a few heads.

The second, ENAFIN 2024, covers only companies with 6 or more people in cities of 50,000 inhabitants or more [7]. There, 61.4% carried out at least one financial planning action in 2024. Put the other way round: almost 4 out of 10 did none [7][8]. If your company is among them, test 4 on the list at the end is a good place to start.

What does the Churchill and Lewis model say, in short?

Each stage of a small business calls for a different owner and a different back office.

In two lines: in May 1983, Neil Churchill and Virginia Lewis, of the Caruth Institute at Southern Methodist University (SMU), published in Harvard Business Review five stages of the small business: existence, survival, success, take-off and resource maturity [1][2]. They built on earlier models, among them Larry Greiner's (1972) [1][4].

Three ideas from the article are useful to you:

• Not every company has to grow or die. Among the companies they studied, several had stabilized without growing for years, some with very good profits [1].
• Each stage is described with five factors: management style, structure, how formal the systems are, the goals and how involved the owner is [1].
• Size is not just how much it sells. Sites, product lines and how fast its technology changes also count. With two or three sites, management problems are more complex [1].

Which stage is your company in?

Place yourself by signals you can count, not by how it feels: people, sites, legal entities and who closes the month.

Churchill and Lewis do not set how many people correspond to each stage. For a size baseline, Mexico has official cut-offs from the Ministry of Economy, published in 2009 [9]:

But size is not stage: a micro company can be healthy and stable, and a medium-sized one can still depend on the owner for everything [1].

The table below crosses the stages with measurable signals. It is indicative and it is our reading, built on the model's description of each stage [1]:

Two rules from the model to read it properly:

• What matters changes from stage to stage. Cash is critical at the start, manageable at success and central again when growing. The owner's ability to do matters at the beginning; the ability to delegate, later. People, planning and systems gain importance from success to take-off [1].
• A company can be ahead on one factor and behind on another. The authors found cases like that, and they warn that the imbalance causes serious problems even though it is rarely more than one stage [1].

To measure yourself, four thresholds. The first comes from the law; the other three are indicative and our reading:

About the 35 million: in the RESICO, income counts when it is collected; under the general regime, as a rule, from when it is invoiced, delivered or collected, whichever happens first [10]. If you are going to cross that threshold, your back office has to be able to track both. Confirm your case with your accountant: this is not a substitute for tax advice.

Does your company stop when you are not there?

At the existence stage, the owner is the company; the back office's job is to make that stop being true on paper.

What it looks like. You sell, deliver, collect, buy and answer the phone. If you are sick for a week, the company stops for a week.

What the model says. The owner does everything and directly supervises a few employees; the strategy is to stay alive. Systems and formal planning are minimal or non-existent [1]. Many companies do not get past this point: they close when the start-up capital runs out [1].

In Mexico. Among micro companies, 4 out of 10 do not solve the problems that come up in production, according to ENAPROCE [5].

What breaks and what to decide beforehand:

• That the company's money and yours live in separate accounts.
• That every sale and every purchase has its tax receipt, where applicable, and that they are kept in one place.
• That someone other than you can answer "how much did we sell this month?".

Why do you sell and still run short of money?

At the survival stage, cash rules: what you do not know every week, you pay for at the end of the month.

What it looks like. There are already repeat customers and a sales supervisor or a shop foreman. But every payday there is payroll, suppliers and taxes to pay, and customers pay when they can.

What the model says. The problem becomes what comes in against what goes out. Supervisors carry out the owner's orders, and planning, when it exists, is a cash forecast [1]. Many companies stay here for a long time, with profits that are barely enough, until the owner gives up or retires [1].

In Mexico. In ENAFIN 2024, 58.9% of companies said something held back their growth in 2023. The most cited reason was the country's economic situation (53.8%). It was followed by strong competition (45.5%) and lack of government support (43.0%). Financing weighed less: its cost was mentioned by 25.8% and its lack, by 20.7% [7]. What holds you back most comes from outside; what does depend on you is knowing how much cash you have.

What breaks and what to decide beforehand:

• Knowing every week how much you are owed, by whom and since when.
• Knowing which expenses have a valid tax receipt and which do not: an expense without a valid receipt can affect its tax treatment. Confirm it with your accountant.
• A forecast for the next four weeks: what comes in, what goes out and on which day it gets tight.

Grow or stay as you are?

Success forces a decision that many owners do not make on purpose: use the company to grow or to live off it.

What it looks like. The company makes a profit. You have already hired someone for the back office, perhaps a controller. For the first time you can take a vacation without everything falling apart.

What the model says. The stage splits into two paths [1]:

• Stay ("D", for disengagement). The company is healthy and can stay that way for years, with functional managers, a controller and basic finance, sales and production systems. The risk: spending the cash of the good years and being left weak for the bad ones [1].
• Grow ("G", for growth). The owner pulls together the company's cash and credit and risks them to grow. The owner has to make sure the core business stays profitable and hire managers with the coming company in mind [1].

For the growth path, the authors call for installing systems "with attention to forthcoming needs", thinking ahead, and for getting people, planning and systems in place a little before growing [1].

What breaks and what to decide beforehand:

• If you stay: that the back office works without you. A month-end close on a fixed date, collections that do not depend on your memory and a controller who reviews discrepancies instead of keying in data.
• If you grow: that what works today with one accountant and three spreadsheets can handle twice the invoices, customers and legal entities. That decision is made here, not at take-off.

Why does cash get tighter the more you sell?

At take-off, two kinds of owner fail: the one who believes anything is possible and the one who believes they have to know everything themselves.

What it looks like. You open a second plant, a branch in another city, a second legal entity. Large customers arrive who pay on longer terms. Sales go up and cash feels tighter than ever.

What the model says. The key questions are how to grow fast and how to finance it. Does the owner truly delegate, with controls, or abdicate, something that according to them happens often? Systems are strained by growth and have to become more refined and more extensive [1].

In Mexico, credit does not arrive on its own. According to ENAFIN 2024 [7]:

• From 2022 to 2024, 17.1% of the companies that applied for credit had their largest application rejected.
• 72.0% of those that did not get it were affected; the main effect was delaying their expansion or the purchase of machinery (45.0%).

In ENAPROCE 2018, among the reasons for denying credit in the previous two years were having no collateral (18.6%), a poor credit history (17.2%) and not being able to prove income (13.6%) [5]. That third one does depend on your back office: orderly numbers also help when you ask for money.

What breaks and what to decide beforehand:

• Several legal entities, banks and branches that have to reconcile in a single close.
• One deposit that pays three invoices, one invoice paid in two installments, a credit note.
• A manager who can decide with reliable numbers without asking you.

How do you consolidate what you have gained without becoming slow?

At maturity, the challenge is for the numbers to arrive on their own without bureaucracy eating up speed.

What it looks like. There are functional managers, several sites and a committee that meets every month. But the sales report and the finance report do not give the same figure, and someone spends the first week of the month reconciling them.

What the model says. What growth left behind has to be consolidated and controlled without losing the speed of a small company. Budgets, strategic planning and standard costs arrive, and the owner and the company become separate [1]. The danger is a kind of sixth stage, ossification: a company that stops innovating and avoids risk until the market changes [1].

What breaks and what to decide beforehand:

• The same data keyed into two systems, with two versions of the truth.
• Reports that someone still puts together by hand in a spreadsheet.
• A systems decision that is no longer which tool to buy, but how the ones you already have connect.

What does the sequence look like in a Mexican company?

Rotoplas first put its processes and its measurement in order, then rolled out one system in all its countries and on top of it centralized the back office.

Rotoplas, known for its rooftop water tanks, told its story year by year in its 2014 annual report to the Mexican Stock Exchange [11]. What follows comes from there, except where another source is indicated. It is an illustrative reading: the stage column is ours, and the company does not use the model or classify itself that way.

One nuance before the table: the case illustrates take-off and maturity, not the early stages. In 1995 Rotoplas already had 500 employees [11], more than the ceiling for a medium-sized company in the official cut-offs [9]. Of its years of existence and survival, the report gives no figures.

(A 2007 article in Expansión places the first plastic water tank in 1984 [15]; here we follow the year given by the company and El CEO [11][14].)

What it achieved, in figures. As of the date of that report, Rotoplas operated in 12 countries, with 24 plants and 5 distribution centers. In 2014 it sold 6,551.8 million pesos [11].

What it cost them, in what can be seen. Twice it gave up part of the company to grow or institutionalize: 20% to the fund in 2003 and, at the IPO, 144.2 million shares representing 29.7% [11]. What the report does not say is how much it cost to install the ERP, or how hard it was.

Note the order of the back-office steps, according to the report [11]:

1. 2003, processes. It restructures to work by processes that could «seguir soportando su crecimiento» [keep supporting its growth].
2. 2005, measurement. It adopts the balanced scorecard as its strategic planning system.
3. 2008, one system. It starts installing the same ERP in all its countries, and that standardizes its processes.
4. 2009, centralized back office. On that basis it creates a shared services center for finance, accounting, collections and customer and distributor service.

The lesson, in our reading: first you put the processes in order and then you buy the system. And the common system is what later lets you centralize the back office of many sites in one place. The report does not say what systems existed before 2008; only that the common one arrived that year.

The next article in the series covers the scorecard that Rotoplas adopted in 2005.

What changes in Mexico with the CFDI?

In Mexico, much of the raw material for your information system already exists in your CFDI; the challenge is keeping it in order.

On the 1983 map, a company at stage 1 may have no systems. In Mexico, tax law adds something.

When the law requires tax receipts to be issued, they are issued as digital documents through the SAT, Mexico's tax authority, and whoever buys, receives a service or pays in installments or on credit must request theirs [16]. In other words: the CFDI (electronic invoices) that must be issued or received already exist in digital form [16].

That changes the reading of the model:

• The good news: a Mexican SME has, from early on, much of the raw material of an information system: what it sells and what it buys with a tax receipt.
• The risk: if that data is not in order, discrepancies between what your CFDI, your bank and your accounting say are discovered late, at the close or during an audit.
• What the CFDI does not include: your inventory, your costs or the aging of your receivables. In sales to the general public, the receipt carries a generic code instead of the customer's RFC (taxpayer ID) [16]. And cash still carries weight: in ENAFIN 2024, 79.8% of companies accepted it in a typical month of 2023 [7].

That is why, in Mexico, the model's "formal systems" can start before the ERP: with your own CFDI complete and reconciled against your bank. This is our reading and is not a substitute for your accountant's advice.

What breaks in your back office at each stage?

Each stage breaks something different; the systems decision is best made a little before growing, not when things are already falling apart.

The model says systems go from minimal to basic, from basic to strained by growth and, in the end, to extensive [1]. What that means in a Mexican company is our reading:

Which mistakes cost the most?

Almost all of these mistakes are made in one stage and paid for in the next; six are described by the model and one is our reading.

1. Growing without backup cash. It is the omnipotence syndrome [1]. Consequence: according to the authors, a company that fails at take-off can fall back to earlier stages or disappear [1].
2. Abdicating instead of delegating. Letting go without controls [1]. Consequence, in our reading: errors that nobody sees in time, and the owner ends up taking everything back.
3. Not letting go of anything. It is the omniscience syndrome [1]. Consequence: the authors point out that the inability to stop doing and start managing explains the downfall of many companies as they grow [1].
4. Spending the cash of the good years. The risk of the stay-as-you-are path [1]. Consequence, in our reading: the company reaches the bad years weak.
5. Carrying the tax habits of survival into the growth stage. The authors warn that delaying tax payments seems vital at first, but later it can distort the numbers and consume management time [1]. Consequence: decisions are made with figures that do not reflect reality.
6. Skipping the school of the early stages. In franchises, the lack of that experience can do damage when the owner moves on to running several units [1]. Consequence: problems the company did not learn to solve when it was small.
7. Putting systems in place after growing. Our reading, based on the model [1]. Consequence: each new branch or legal entity adds its own way of keeping the accounts, and the close gets longer.

What should you take to Monday's meeting?

Ten tests that can be measured in a week tell you where your bottleneck is and where to start.

The list is our own tool, not part of the model. A "no" does not say which stage your company is in: it says what was left pending from that stage. Answer it with data, not from memory, and fill in the last three columns at the meeting.

How to read it: each stage has two tests. Go through them from 1 to 10: the first stage with a "no" is your bottleneck, and that is where you start. What was not solved earlier is carried into the next one.

And two steps for after the meeting:

1. Ask your accountant for the number of CFDI from last month, issued and received, and how many had to be chased.
2. Check a deposit at random from last month. If nobody knows which invoice it belongs to, you already know where to start.

Key takeaways

Three questions for your next executive committee:

1. Do we want to grow or stay as we are, and did we decide it on purpose or by inertia?
2. If we opened another branch or legal entity tomorrow, would the following month's close come out on time?
3. Which part of our back office depends today on a single person?

What does Hábil do?

What breaks most between survival and take-off, CFDI against the bank and several legal entities, is what Cuadre addresses.

Hábil is a Mexican engineering consultancy that has been building critical systems since 2006 [17].

Cuadre downloads from the SAT the CFDI your company issues and receives, including payment complements. It continuously reviews what your suppliers invoice you and notifies you on WhatsApp when a new one arrives. On that basis it reconciles on three fronts:

• payment complements against your invoices;
• what the SAT says against what your accounting system recorded;
• your bank deposits against your CFDI.

What does not reconcile is left in plain view so your team can resolve it before the close. You authorize the connection, and before we start we give you in writing how it connects and how it is removed.

Cuadre works with all of your company's RFCs. It does not replace your accountant; it leaves them the discrepancies instead of the whole month.

Learn about Cuadre at habil.mx/en/cuadre. For the details, read Why bank reconciliation does not add up at a Mexican company.

Request your first conversation, at no cost. We review with you how you close your month today and show you Cuadre with your own tax receipts. Write to us on WhatsApp.

Frequently asked questions

Is the model still valid if it was published in 1983?

It works as a map, with limits. The authors tested it with 83 responses from owners and managers who were taking part in a program for small businesses and had already read Greiner's article [1]: a small sample, predisposed to the framework. It is still cited in Mexican academia, for example in a 2018 study of micro companies in Culiacán [3]. And it predates the CFDI [16].

Do I need an ERP to move to the next stage?

Not at first: the first step is to have your CFDI complete and reconciled with the bank every month. The model asks for systems to arrive a little before growth [1]. An indicative rule, our reading: think about an ERP when two of these apply: you have two or more legal entities that must close together; you handle inventory or production that has to be costed; you operate in two or more sites; or you key the same data into two systems.

Did Hábil work with Rotoplas?

No. The case uses only public sources, cited in the references.

Sources consulted on October 22, 2026. The paraphrases of the Churchill and Lewis article are our own; the only direct quotation is a brief excerpt, in English, from the original.