The Technology Expense Management Buying Guide: How to Choose a TEM Platform or Managed Program

Technology spend crosses a line somewhere past a handful of locations and a few dozen services. Below the line, one person can say what the company pays for, which of it is still used, and whether the invoices match the contracts. Above it, nobody can. Billing for canceled circuits continues for years. A rate that was supposed to drop at renewal never does. Mobility pools carry devices that left with the employees who held them. Cloud and software subscriptions accumulate under corporate cards. Finance reconciles by exhaustion, and every month the estate is a little less knowable than the month before.

Technology expense management exists for exactly this, and the category is bought badly more often than it is bought well, because the proposals all describe the same outcome and the programs behind them differ enormously. This guide is the evaluation we run when we advise on expense and asset management, written out so you can run it with an advisor or entirely on your own.

Decide what you are buying

Three different things are sold under the same name, and the first decision is which one you need.

An audit is a one-time recovery project. The provider takes a defined window of invoices, contracts, and carrier records, finds the errors, files the disputes, and hands back credits and a cleaned inventory. It ends. Pricing is usually a share of what is recovered or a fixed project fee.

A platform is software your team operates. It loads invoices, holds the inventory, flags variances, and produces the reports. The provider supplies tooling and support; your people supply the hours. A platform without hours behind it becomes expensive shelfware with dashboards.

A managed program supplies the hours too. The provider loads and validates every invoice, files and chases disputes, processes orders and changes, and keeps the inventory true as the estate changes. Co-managed versions split the work: the provider does the processing, your team approves and decides.

The right sequence for most companies is audit first, program second. The audit recovers what is already owed and produces the clean baseline. The program keeps it clean. Starting a program without the audit means paying to manage waste faithfully. Buying the audit alone means watching the estate drift back over the next two years. Be honest about internal capacity when choosing between a platform and a program, because the capacity question decides whether the value stays theoretical.

Scope the estate before the first demo

Proposals are only comparable when every provider is pricing the same estate, and most buyers cannot describe theirs when the process starts. Build the description first. It is the same exercise that opens the software buying guide: start from the money rather than from memory.

Pull twelve months of accounts payable for every carrier, mobility provider, cloud platform, and software vendor, plus the corporate card statements where the unmentioned subscriptions live. From that, capture what a provider needs to price the work:

• Domains. Fixed telecom and data circuits, mobility, cloud infrastructure, software subscriptions, and whether hardware and asset lifecycle belong in scope. Each domain needs different data sources and different expertise, and the provider's depth in each is the thing to test.
• Volume. Invoice count per month, vendor count, line and device counts, and the number of legal entities and cost centers the spend has to be allocated across.
• Complexity. Currencies, countries, the carriers involved, and how many invoices arrive as paper or portal downloads rather than electronic feeds.
• Systems. The accounts payable platform, the general ledger structure, the ITSM or ticketing tool where orders originate, and the HR system that says who still works here.

This document does two jobs. It lets every provider price the same thing, and it exposes the first savings before a contract is signed, because the act of listing what the company pays for surfaces the services nobody can explain.

The inventory is the product

Every promise a provider makes rests on the inventory: the list of every circuit, line, device, subscription, and license the company holds, with the contract terms, the rate, the location, and the owner attached. A wrong inventory produces wrong disputes, wrong allocations, and confident reports about a fiction. So the questions that matter most are about how the inventory gets built and how it stays true.

Ask how the initial build works. The credible answer combines several sources: invoice parsing, carrier service records and portal exports, contract review, and discovery against the network and the mobile device management platform. A build that relies on invoices alone inherits every error the invoices contain. Ask how long the build takes for an estate like yours, what it requires from your team in hours and in access, and what the acceptance test looks like at the end. An inventory should be signed off, not assumed.

Then ask how it stays accurate. The honest mechanism is order management: when every add, move, change, and disconnect flows through the provider, the inventory updates as the estate changes. When orders bypass the provider, the inventory drifts, and within a year the program is a report generator. Decide before signing whether your team will route orders through the platform, because a program that depends on it and does not get it fails quietly.

Finally, confirm in writing that the inventory is yours. It is the most valuable thing the engagement produces, and it should leave with you in a usable format on the day the contract ends.

Who files the disputes

Recovering billing errors is the core promise, and it is a promise about labor. Software flags a variance. A person files the dispute, supplies the evidence, follows up when the carrier ignores it, escalates when the carrier denies it, and confirms the credit actually landed on a later invoice. Ask precisely who does each of those steps under the proposal in front of you.

Then ask for the evidence behind the recovery claims. Which error types does the provider catch: rates above contract, charges for disconnected services, taxes and surcharges applied wrongly, late fees caused by slow processing, usage outside plan. What recovery rates do comparable clients see, measured how, against what baseline, over what period. A provider that answers precisely has a process. One that answers with an industry average has a brochure.

Two adjacent capabilities decide whether the program closes the loop with finance. General ledger coding and allocation turn a validated invoice into a cost that lands on the right site, department, and entity without a spreadsheet in the middle. Payment, where offered, moves the approved invoice into accounts payable or pays it on your behalf, which ends the late fees that slow internal approval creates. Neither is mandatory, but the integrations that support them are the difference between a program finance relies on and one it tolerates.

Pricing models, and how each bends as spend falls

Technology expense management prices along a small number of models, and the model matters more than the rate, because a good program shrinks the very spend the fee is based on.

• Percentage of managed spend. A rate against the invoices under management. The fee falls as the provider eliminates spend, which aligns incentives, and rises if the estate grows. It is hard to forecast, and it can quietly reward leaving expensive services in place. Ask what happens to the rate as spend declines, and whether there is a floor.
• Fixed per unit. Per invoice, per line, per device, per user, or per entity. Predictable, and often cheaper at scale. It does not shrink when the estate shrinks unless the contract says so, and minimums can turn a fixed price into a floor you no longer need.
• Platform subscription plus services. A software fee for the tooling and a separate fee for managed work. Transparent about what the hours cost, and prone to scope fees when the work turns out to be larger than the sales estimate.
• Savings share. A share of recovered credits and realized reductions, often layered on an audit or a lower base fee. Attractive because the provider is paid from money they found, and worth reading closely for how savings are measured and for how long the share runs.

Implementation and the inventory build almost always carry a one-time fee, and it is the most negotiable line in the proposal. Run three years of your own invoice counts and a realistic spend reduction through each candidate's model before comparing headline rates. A percentage proposal that looks cheap in year one can cost more than a fixed one by year three, or the reverse, and the ranking depends on numbers only you have.

Demand the method behind the savings claim

Every provider in this category quotes savings, and the number is only meaningful with its method attached. When a proposal says a client saved a large share of spend, ask which client, at what starting state, measured against what baseline, over what period, and whether the number counts one-time credits, recurring reductions, or both. Recovered credits are real but do not repeat. Recurring reductions are the number that matters for your budget.

Then ask for references that resemble you: similar domains in scope, similar invoice volume, similar entity structure, and a program at least a year old. A reference from the first six months of a program describes the audit. A reference from year two describes whether the inventory held. Ask the reference two questions above all others: does the inventory match reality today, and who chases the disputes.

The quote comparison method applies here with one adjustment. In this category the proposal's own savings projection belongs in the comparison, discounted by how well the provider defended it. A cheaper program that recovers less is the more expensive option.

The contract terms that protect you

The terms below cost nothing to include at signature and a great deal to obtain afterward. Every one of them should be in the agreement before it is signed.

• Implementation milestones with acceptance. A dated plan for the inventory build with a defined acceptance test, and fees tied to milestones rather than to the signature.
• Processing service levels. Days from invoice receipt to loading, validation, and approval routing, with a remedy when they slip. Slow processing creates the late fees the program was supposed to end.
• Dispute reporting. Disputes filed, disputes won, credits confirmed on subsequent invoices, and the aging of what is still open, reported monthly in a form finance can audit.
• Scope and fee flexibility. The right to reduce units, entities, or domains as the estate shrinks, with the fee following, and a defined price for adding scope so a merger or a new domain does not reopen the whole agreement.
• Price protection. A cap on annual increases, applied to renewals and to services added later.
• Data ownership and exit. The inventory, invoice images, contract library, and dispute history are yours, exported in a usable format at no charge at the end of the term, with a defined transition period.
• Term and renewal. A term short enough to test the program, a renewal notice deadline you can meet, and a termination right for sustained service failure.
• Security and access. Carrier portal credentials, invoice data, and employee records under the provider's control, with the security commitments and breach notification terms that access deserves.

The renewal playbook covers the calendar and evidence discipline that make the price protection and notice terms enforceable when the renewal arrives.

Scope the program where the money leaks today

Fixed telecom is the historical core of this category, and for many companies it is no longer where the waste lives. The fastest-growing leakage now sits in mobility pools, where devices outlive the employees who carried them, and in cloud and software subscriptions, where a bill that grows by default and seats provisioned in a hiring year go unchallenged because nobody owns them.

Providers vary widely in depth across these domains. A telecom-rooted program may handle mobility well and cloud superficially. A cloud cost specialist may not touch a circuit invoice. Test depth per domain rather than accepting logo-slide breadth: ask to see the actual workflow for a mobility line disconnect, a cloud reserved-capacity review, and a software seat reclaim, with the reports each produces. If one domain carries most of your leakage, a specialist alongside a narrower program can beat a single provider claiming everything.

Red flags worth slowing down for

• The proposal arrives before anyone has asked for your invoice counts, domains, or entity structure.
• Savings are quoted as an average without a method, a baseline, or a client that resembles you.
• Nobody can say precisely who files disputes and who chases them.
• The inventory is built from invoices alone, with no acceptance test.
• Order management is optional in the sales conversation and mandatory for the inventory to hold.
• The fee does not fall as the estate shrinks, or a minimum keeps it from falling.
• The inventory and dispute history are not clearly yours at exit.
• Every domain is covered, and the demo of any one of them is a slide.

Where an advisor fits

All of the above is runnable on your own, and this guide exists to make that practical. What an advisor adds is the market work that a company buying expense management once every several years cannot do from the inside: we help you scope the estate and decide which shape of program fits, then run the providers that match it across the whole market against your actual domains, entities, and invoice volumes, with pricing and terms benchmarked against what companies like yours actually sign. Because we also source the underlying services, the loop closes: the inventory findings become the evidence position for the renewals that matter, and every circuit or contract sourced through the process arrives documented. You get 3 to 5 recommended suppliers with our reasoning on each, proposals normalized so percentage and fixed pricing compare honestly on net cost, and contract terms checked before you sign. It costs you nothing, we have no stake in which supplier wins, and you sign directly with the one you choose.

If finance has asked what the company actually pays for and nobody could answer, if invoices arrive by the dozen and get paid on trust, or if a renewal is coming for a carrier agreement nobody has read, a thirty-minute conversation before the next proposal lands will save you from managing waste faithfully for three more years.