InsightsPoint of View

The Business Logic We Outsource

Owning the records does not mean controlling the decisions

A business owner reviewing a customer order while an independent delivery partner waits beside a protected fixture crate.
By
Cam Myers
Content type
Point of View
Reading time
7 min read
Status
Author’s perspective
Topics
Ownership & Economics

In brief

  • Data possession, authority to change a rule, and practical ability to move an operation are different forms of control.
  • Provider dependence can be an excellent purchase when its value and consequences are understood.
  • A usable transition requires operating knowledge and continuity as well as an export of records.

The argument

You can own your customer list, pay for every software account, and still have surprisingly little control over how your business makes its next commercial decision.

An advertising platform decides which prospective customers deserve a bid. A booking system decides which requests can become orders. A pricing tool recommends what to charge. Each may work well. But somewhere inside those arrangements are rules about what counts as valuable, what information matters, and which tradeoffs are acceptable.

Those rules are part of the business model in operation. Outsourcing them deserves the same attention as outsourcing the work they direct.

The issue is not whether a provider knows more about software than you do. It probably does. The issue is which decisions it can make, which decisions you can change, and whether you still understand the connection between those decisions and the money the business keeps.

The decision underneath

Can the business still direct the logic through which its resources become commercial action?

The decision hidden inside the service

Consider a hypothetical regional equipment-rental business in the Midwest. It uses external software for advertising, quoting, booking, and scheduling. Management owns the accounts and can export customer records. That sounds like a reasonable degree of control.

Suppose its booking software accepts an order whenever equipment is available and the payment succeeds. That is convenient for ordinary jobs. But the business also serves sites where delivery access, distance, or installation conditions require a different price or a conversation before acceptance.

The customer can enter an address and describe the site. Those facts appear correctly in the exported record. Yet if the software cannot use them to change eligibility or require review, management has little control over the commitment made at checkout. Owning the information has not given the business authority over the rule applied to it.

The question is whether the commercial policy can be expressed where the commitment occurs. A configurable service-area rule may be sufficient. An unusual request might remain a quote until someone confirms the conditions. A separate application could supply the decision if the existing platform supports the connection. If none of those arrangements is practical, the limitation belongs in the investment comparison.

The rental business need not replace its whole booking system to regain control over one consequential choice. Nor should it add custom software if a standard platform already handles the requirement economically. It needs to establish whether the arrangement can carry the customer promise it intends to make.

Advertising exposes the same distinction at an earlier stage. Google's documentation explains that value-based bidding uses advertiser-supplied conversion values to guide bids.[1] A business that supplies gross transaction value has instructed a different objective from one that supplies a dependable estimate of contribution. Choosing the signal is part of the commercial policy delegated to the platform.

Control has several parts

Possessing the records is one form of control. Being able to change the rule applied to those records is another. Being able to move the operation somewhere else is a third.

These can come apart. An export might preserve customers and orders without preserving the logic that determines eligibility, price, priority, or exceptions. A custom application might provide source code while depending on a specialist whom the business cannot readily replace. An account administrator might be able to change a setting without understanding the financial consequences.

For an important capability, management needs to know where the policy resides and how it is revised. Suppose the rental company wants to prioritize longer commitments in a particular service area. Is that an editable business rule, a change requiring the vendor's development team, or something the platform cannot represent? The answer changes the cost and speed of pursuing the strategy.

Authority also needs to be distinguished from advice. A tool that recommends a price leaves someone to make the decision. A tool that publishes the price, accepts the booking, and reserves equipment has been given operating authority. The appropriate oversight depends on the consequences and how readily an error can be corrected.

This does not mean managers must inspect every model or approve every transaction. They need enough visibility to judge whether the capability serves its purpose, identify consequential failures, and change direction. That may come from documented rules, transaction records, comparison tests, exception reports, or access to someone who can explain and modify the arrangement.

Without that capability, retaining nominal ownership can leave the business dependent on decisions it cannot meaningfully govern.

Dependence can be a good purchase

There are strong reasons to rely on external platforms. Their scale, specialist knowledge, distribution, maintenance, and ongoing development may produce capabilities the individual business could not justify building.

Trying to remove every dependency can create a larger burden than accepting one. A company that builds its own system inherits staffing, support, security, recovery, and integration responsibilities. Paying a provider to absorb those obligations may be an excellent use of money.

The Australian Government's architecture guidance makes a useful distinction: provider dependence should be understood and weighed against the value gained; maximum portability is not automatically the best outcome.[2] The same economic discipline applies outside government. Independence has a cost, and its value depends on what is at stake.

Return to the rental business. It might sensibly buy a standard scheduling system while maintaining its service-area definitions and commercial priorities in a form it can inspect and revise. Or a specialist platform may already express those rules well enough that adding another software layer would increase cost without improving a consequential decision.

The useful boundary is specific. Which parts distinguish the business commercially? Which require frequent changes? Which failures would interrupt operations? Which obligations would management rather pay a capable provider to carry?

A business should be able to explain why it accepts a material dependency. A deliberate choice can be revisited when prices, strategy, capability, or operating conditions change.

Preserve the ability to change direction

An exit clause is useful, but it does not move an operation. A usable transition may require customer and transaction history, business-rule definitions, identifiers, integration knowledge, account access, and people who understand how the pieces fit.

That is why a successful data export is only part of an exit test. Can another system interpret the records? Can the business reconstruct the decisions? Can essential work continue during the change? The answer need not be an immediate switch to a competing platform. A slower but workable transition may be entirely acceptable if it matches the consequence of disruption.

The same questions belong in an acquisition or major investment review. A capability may look efficient because crucial knowledge remains with a vendor or one employee. Its continuing value depends partly on whether that knowledge can be retained, transferred, or replaced at a sensible cost.

AI makes this distinction increasingly relevant wherever tools move from producing suggestions to taking actions. More capable software expands what can be delegated. It also expands the importance of specifying whose objective it pursues and where its authority ends.

Owners do not need to possess every line of code. They need a deliberate arrangement for the decisions that matter: enough understanding to assess them, enough authority to change them, and enough continuity to keep the business operating when an arrangement stops serving its purpose.

The strategic question is not simply whether the business owns its data or rents its software. It is whether the business can still direct the logic through which those resources become commercial action.

Sources

  1. Google Ads Help. About conversion values.
  2. Australian Government Architecture. Guide to managing lock in portability and exit planning.

For the leadership team

In practice

  • Locate the policy behind consequential commitments, including eligibility, pricing, and exceptions.
  • Evaluate the cost of changing that policy alongside the capability purchased from a provider.
  • Test whether another arrangement could interpret the records and reconstruct essential decisions.

Questions worth asking

  1. Which commercial rules can management inspect and revise where a customer commitment is made?
  2. What would the business need to preserve if this provider or specialist became unavailable?

Weighing this decision now? Tell us what you are considering and we will say plainly whether we can help.

Discuss your business