Principles

When a company has a problem, the solution often sought is new software.

Sales are not visible enough – let’s implement a new CRM.

Projects are running late – we need a better project management tool.

Information is not flowing – let’s add a new communication platform.

Reporting takes too long – let’s buy an analytics solution.

Customer service cannot keep up – let’s add a chatbot.

People are not collaborating enough – let’s implement yet another shared workspace.

Buying new software feels like a concrete management decision. The problem has been identified, the solution selected, and the project launched.

But software does not automatically create clarity, accountability, or effective ways of working within a company.

It digitizes the logic that already exists within the company.

If the process is good, software can make it faster and more reliable.

If the process is flawed, software can make the flaw spread faster, cost more, and become much harder to fix.

Software does not solve a problem the company itself does not understand

Management often begins by selecting a tool before clearly defining the problem to be solved.

They say they need a better CRM.

But what does “better” mean?

Are salespeople failing to enter information?

Are the stages of the sales process unclear?

Does the company not know which leads are valuable?

Can the manager not see which deals are actually progressing?

Does preparing proposals take too long?

Does the existing software lack the required capability, or has the company simply failed to configure it properly?

These are different problems and require different solutions.

A new CRM will not fix an unclear sales process. It will simply give the unclear process a new user interface.

The same applies to every other type of software.

If a company cannot say which specific task, decision, or outcome needs to change, it cannot know whether the new software solved the problem.

Implementing the software becomes an end in itself.

Software does not replace the decision about how the company should operate

Before configuring a system, the company must decide:

• what work actually needs to be done;
• who is responsible for it;
• what information is required;
• who makes which decision;
• when work moves to the next stage;
• what outcome indicates that the work is complete;
• which exceptions are allowed;
• which activities should no longer be performed.

Software cannot make these decisions for the company.

It may offer a default workflow, but that workflow is based on the software vendor’s idea of a typical customer.

It may not reflect the business model, customers, or accountability structure of a particular company.

If the company has not thought through its own way of working, the software will begin to shape it.

The sales process becomes whatever the CRM allows.

Project management becomes moving tasks from one column to another.

Performance becomes whatever the standard report can display.

The company may thus begin adapting its work to the software instead of having the software support the best way for the company to work.

A new tool often conceals a management problem

Buying software is easier for a manager than clarifying accountability.

If information is not flowing, the company can buy a new communication platform.

But the problem may be that people do not know what information to share, with whom, or when.

If tasks are left undone, the company can implement a new project management system.

But the problem may be unclear priorities, missing decisions, or the fact that one person has been given more responsibility than they can handle.

If sales results are inconsistent, the company can add sales automation.

But the problem may be the wrong target market, a weak value proposition, or insufficient competence on the part of the salesperson.

When a management problem is addressed with software, the management problem does not disappear.

The company now has the same problem along with a new monthly fee, configuration work, and training needs.

Every department optimizes its own work, and the company loses sight of the whole

The number of software tools often grows department by department.

Marketing chooses the tools that suit it.

Sales uses its own system.

Project management adopts another one.

Customer support adds its own platform.

Finance uses a different dataset.

Each choice may make sense when viewed in isolation.

The problem emerges between them.

Customer information is in one system, the contract in another, the project in a third, and the invoice in a fourth. The same data is entered repeatedly. The fields and definitions used by the systems do not match. In one system, the customer is active; in another, the relationship has ended; and in a third, the customer does not exist at all.

Integrations are then built.

If an integration does not cover every exception, a manual check is added. If people do not trust the system, they create their own spreadsheets.

The company may ultimately have many good software products but no single end-to-end workflow.

Each department is more efficient from its own perspective.

The company as a whole spends more time resolving confusion between systems.

More software creates more manual work when systems do not fit together

The purpose of software should be to reduce manual work.

A poorly designed software environment does the opposite.

People:

• copy information from one system to another;
• export data to spreadsheets;
• compare different reports;
• fix integration errors;
• verify which information is correct;
• enter the same customer data multiple times;
• send screenshots because someone else lacks access;
• maintain their own personal backup list;
• ask colleagues for information that should be available in the system.

None of these activities creates value for the customer.

They exist only because the company’s software does not form a coherent whole.

The more systems are added, the more connections are created between them.

Every new connection requires configuration, maintenance, access management, data mapping, and error handling.

At some point, the company is no longer focused solely on its core business.

It is managing a software ecosystem of its own making.

The true cost of software is not the monthly fee

Software selection often focuses on comparing license fees.

But the license may represent only a small part of the total cost.

The true cost of software includes:

• selection and procurement;
• configuration;
• cleaning up data;
• migration;
• integrations;
• employee training;
• changing work processes;
• user support;
• access rights management;
• security risks;
• report customization;
• fixing errors;
• running the old system in parallel;
• the time spent by people who must adapt to the new solution.

The company must also account for the possibility that the software may never truly work in practice.

People use only a small fraction of its features. Some employees remain on the old system. Someone creates a separate spreadsheet on the side. Management does not trust the reports and continues to request manually prepared summaries.

The company pays for the software, but the work happens elsewhere.

An inexpensive license can become a very costly solution if it does not fit the company’s process or people do not adopt it.

Unused software is not a technical problem

When employees do not use a new system, the conclusion is often that people resist change.

Sometimes they do.

But the reason for the resistance may be entirely rational.

The software may:

• add new steps to the work;
• request information that no one uses later;
• duplicate an existing system;
• be too slow for the actual work;
• fail to support important exceptions;
• provide less value to the employee than to the manager;
• require a level of precision that cannot be achieved at an earlier stage of the process;
• make a simple task more complicated.

If someone must do their actual work in one place and then enter the same information into another system for management purposes, it is hardly surprising that the data remains incomplete.

Employees do not adopt software simply because the company pays for it.

They use it when the software helps them do their work better, faster, or with less risk of error.

Adoption does not begin with training.

It begins with a suitable work process and clear value for the user.

Waste can be automated too

Automation always sounds efficient.

If an activity takes a lot of time, automating it seems like a logical step.

But before automating it, the company must ask whether the activity is needed at all.

If a company automates the production of an unnecessary report, the report does not become more necessary.

If a sales message aimed at the wrong customer is automated, an irrelevant offer simply reaches more people.

If a flawed process is embedded in software, changing it later becomes more difficult.

People can adapt a manual process when necessary. Logic built into a system requires analysis, configuration, development, and testing to change.

Three things must therefore happen before automation:

1. Remove activities that do not create enough value.
2. Simplify the necessary process.
3. Automate only the stable and repetitive part.

Otherwise, the company is investing money in doing the wrong things more efficiently than before.

Artificial intelligence does not improve an unclear company either

Artificial intelligence now makes it possible to automate activities that previously required a person.

This creates an enormous opportunity.

But the same principle applies here.

If the input is poor, the process unclear, and the expected outcome undefined, artificial intelligence simply produces more uncertain output faster.

AI can create more sales messages.

But if the company does not know which customers are right for it or what problem it solves, AI scales irrelevant communication.

AI can produce summaries.

But if decision-makers do not know what information they need, the result is simply more summaries.

AI can automate customer support.

But if customer questions arise from a flawed product, automation may hide the problem from management.

Artificial intelligence does not reduce the need to understand the company.

It makes that understanding more important, because a poorly chosen activity can now be performed much faster and at a far greater scale.

Custom software does not automatically solve a unique need

A company may conclude that no off-the-shelf software suits it and that it needs its own system.

Sometimes that is the right decision.

If the company’s operating logic is an important part of its competitive advantage, custom software can create enormous value.

But custom software quickly becomes expensive if the company has not understood its needs clearly enough before development begins.

It then starts making decisions during development about:

• how the process actually works;
• who is responsible for what;
• what data is required;
• which exceptions to support;
• which reports are important;
• how to measure the outcome.

A developer can build what is described to them.

They cannot decide on the company’s behalf what its business logic should be.

If the specification reflects the existing confusion, the company will ultimately get expensive custom software that hard-codes that same confusion.

Building custom software makes sense when the company understands both the problem to be solved and the operating logic the system must support well enough.

Not as a way to avoid uncomfortable management decisions.

Excel is not always the problem, and a new platform is not always the solution

Companies sometimes feel embarrassed by spreadsheets and manual work because they seem unprofessional.

But a simple spreadsheet can be exactly the right tool at an early stage.

If the process is still changing, volumes are low, and the company is learning what information it actually needs, a flexible solution may be better than a large system.

A problem arises when:

• a critical activity depends on one person’s personal spreadsheet;
• data is copied manually to multiple places;
• there is no single source of truth;
• the impact of errors becomes significant;
• the workload exceeds the capacity for manual management;
• the process is stable enough, but the company keeps postponing automation.

The quality of a tool is not determined by its technical complexity.

It is determined by how well it fits the work, volume, risk, and stage of the company’s development.

Software that is too complex can be just as wrong as a spreadsheet that is too simple.

How can you tell that a company has too much software?

The number of software tools is not a problem in itself.

The problem is the total complexity they create.

Warning signs include:

• no one knows all the systems in use;
• different departments use different applications for the same information;
• the same customer’s data differs across several places;
• employees must move information between systems manually;
• important work still takes place outside the official software;
• management does not trust reports generated by the systems;
• the company pays for the same functionality in multiple tools;
• license owners have left the company;
• some systems are used by only one person;
• setting up access for a new employee is a project in its own right;
• fixing integrations takes more time than developing the core business;
• new tools are added, but old ones are not retired;
• people do not know which system is the official source of truth;
• software notifications create more disruption than value.

In this situation, the company probably does not need another piece of software.

It first needs clarity in its existing environment.

Seven questions before buying the next piece of software

Before selecting a new tool, management should answer at least seven questions.

1. What specific problem are we solving?
2. How is this problem harming the company’s performance today?
3. Is the root cause in the software, the process, accountability, competence, or the choice of work itself?
4. Which activity can we stop or simplify before adding software?
5. Which existing system might already meet the same need?
6. What measurable outcome must improve after implementation?
7. Which old tool, manual task, or stage of work will the new solution replace?

The final question is particularly important.

If new software does not replace anything, reduce work, or clearly improve an outcome, it will most likely simply add another layer to the company.

The right sequence is problem, process, accountability, and only then software

The right starting point for a software project is not a product comparison.

First, the company must make visible how work actually happens.

Not how the process manual says it happens, but:

• where the work originates;
• who does what;
• what information they need;
• where the work waits;
• where errors occur;
• what is done repeatedly;
• which decisions have not been clearly assigned to anyone;
• which activities create no value for the customer or the company.

Unnecessary work must then be removed.

The necessary process must be simplified.

Accountability and decision-making authority must be established.

The data must be put in order.

Only then can the company select software that supports the desired way of working.

Otherwise, it is buying a technical solution to an organizational problem.

Good software makes the company less dependent on software

This may sound contradictory.

But a good system does not force people to spend their time managing it.

It works in the background, reduces manual work, keeps information unambiguous, and helps the right person make the right decision at the right time.

People do not have to remember which of five systems to enter something into.

Management does not have to request separate reports.

Customer information is not lost during handovers.

Repetitive work does not require a new person each time.

The value of software does not lie in the number of features.

Its value lies in how much unnecessary work, error, waiting, and confusion it removes from the company.

A company does not improve simply because it has more technology.

It improves when technology helps the right person do the right work with less cost, risk, and time.

More software can create the impression that the company is progressing.

But if every new tool adds a new layer of data, a new workflow, and a new coordination burden, the company is not digitizing its capabilities.

It is digitizing its confusion.

Mikk OrglaanChalleng.ist