Access Before Competition

Why finding the notice is not the same as reaching the market

The tender is public. The deadline is three weeks away. The documents explain the work in 74 pages, refer to five appendices, and require proof of insurance, prior performance, financial capacity, registration, cybersecurity controls, and a pricing schedule in the buyer's preferred format. Somewhere in the package is a contract a capable small firm could perform.

Is that firm competing?

Not yet. First it must decide whether the opportunity is real, whether the requirements are proportional, whether it can assemble the evidence, whether the buyer will recognize its experience, whether the cash cycle is survivable, and whether the bid is worth several evenings of unpaid work. The competition described in the notice begins only after this private tournament has been completed.

This is the difference between visibility and access. A market can publish every opportunity and still exclude firms through the cost of becoming legible to the buyer.

The market before the bid

Public competition is usually measured after a firm submits. The more consequential story begins earlier, in the sequence through which a business notices an opportunity, interprets it, prices the pursuit, and decides whether entering the market is rational.

Public procurement is often discussed through award totals. In fiscal 2024, the U.S. Small Business Administration reported more than $183 billion in federal prime contracts awarded to small businesses. That figure demonstrates scale and policy intent. It does not tell us how many plausible suppliers never entered, how concentrated awards were, or how much capacity was consumed by unsuccessful pursuit.

The global evidence makes the participation gap visible. A World Bank analysis of more than 90,000 firms found that small and medium-size enterprises were, on average, 12% less likely than larger firms to participate in public procurement even after accounting for age, productivity, and sector. Earlier OECD research identified information quality, contract size, administrative burden, and disproportionate qualification requirements among the most common obstacles.

These are pre-competitive barriers. They determine who arrives at the starting line.

The policy response is frequently to improve discovery: consolidate portals, send alerts, publish forecasts, hold matchmaking events. These measures matter. A supplier cannot pursue what it cannot see. But a notice is not a pathway. Better discovery may even increase the burden if it gives owners more poor-fit opportunities to screen.

Consider an illustrative composite: a twelve-person facilities-maintenance company sees a municipal contract covering six buildings. The work resembles jobs it already performs. The owner initially reads the opportunity as growth.

Then the hidden questions arrive. The municipality pays in 45 days, but payroll is biweekly. A performance bond may be required. The insurance limit exceeds the firm's current policy. Two service categories are outside its core work. The buyer asks for three comparable government references, while the firm's strongest work has been for private property managers. The contract is winnable only if the company teams with another supplier—and there is no directory of partners willing to share responsibility on these terms.

The opportunity is “open.” Yet the owner is making a capital-allocation decision with incomplete information. The cost of bidding includes document production, partner search, legal interpretation, pricing uncertainty, and the revenue work displaced by all four.

This is why match quality matters more than alert volume. An opportunity-intelligence system should not ask only, “Is the firm eligible?” It should ask: Is the scope coherent with current capabilities? Which requirements are hard constraints and which allow equivalent evidence? What must be borrowed, bought, or partnered for? How much cash is exposed? What relationship history exists? Which assumptions could make a good-looking opportunity irrational?

A responsible answer may be “do not bid.” That is not a failed recommendation. It is capacity preserved.

Procurement statistics are strongest at the point where the public system has a record: a notice was issued, bids were received, an award was made, money was obligated. The supplier's decision process begins earlier and leaves little trace.

Picture access as a funnel. A firm must first become aware of a market. It must identify a specific opportunity, understand it, believe it is genuinely contestable, establish eligibility, decide that pursuit is strategically and financially rational, assemble the evidence, submit a compliant response, survive evaluation, negotiate or accept terms, mobilize, perform, invoice, and be paid. Each stage can remove an otherwise capable supplier. Only the last stages are visible in award data.

The World Bank's firm-level procurement research is important because it observes firms, not merely contracts. Across the underlying enterprise surveys, private-firm participation in public procurement was about 18%, and small firms were 12 percentage points less likely to participate than large firms. The gap cannot be explained away simply by saying small firms win smaller contracts. Many do not enter.

An award target can rise while the funnel remains narrow. Existing small contractors may win more, a few mature suppliers may account for a large share, or subcontracting dollars may grow without creating new prime relationships. None of these outcomes is worthless. They answer a different question from whether the market is becoming contestable for unfamiliar firms.

A serious access strategy would track stage conversion. How many firms saw the opportunity? How many downloaded the documents? How many attended market engagement? How many asked questions? How many began and abandoned a bid? What requirement drove the abandonment? How many qualified firms declined because of payment or capacity? How many first-time bidders became first-time winners, and how many of those returned?

Most procurement systems cannot answer these questions because they treat nonparticipation as absence rather than data. Yet the owner who reads a notice and decides not to bid may know more about the market's accessibility than the buyer who receives three compliant offers.

Electronic procurement deserves credit for changing the first stages of the funnel. Notices that once appeared in specialized newspapers or physical offices can be searched nationally. Documents can be downloaded without travel. Questions and amendments can be distributed consistently. Submissions can be timestamped and audited.

The World Bank's reviews of e-procurement systems describe lower administrative cost and expanded visibility. In Korea, the KONEPS system became a frequently cited case: after introduction, SME registration for procurement participation doubled. This is meaningful. Reducing the cost of knowing that a market exists is genuine reform.

But registration is not competition, and competition is not delivery. Once discovery becomes easier, the bottleneck moves. Suppliers confront complex classifications, multiple registrations, document formats, digital signatures, security requirements, and the difficulty of distinguishing a live opportunity from a procurement formality. The portal makes the inventory of demand visible; it does not make every item in that inventory usable.

Digital systems can also harden bad process. A contracting officer could once accept a clarification at a counter; a validation rule may now reject the submission. Standardization reduces discretion and the unequal treatment that discretion can permit. It can also make proportional judgment harder. The question is not whether paper or digital is kinder. It is whether the process has been simplified before it is encoded.

This matters for AI-enabled procurement. A model can summarize an 80-page solicitation, extract deadlines, and draft a compliance matrix. That may return real supplier capacity. It may also make a disproportionate requirement easier to endure and therefore less likely to be reformed. Technology can improve participation while preserving the policy choice that created the burden.

The best digital system would generate evidence in both directions. It would show suppliers what the buyer requires and show buyers where qualified firms consistently stop. Access intelligence is not only a matching function. It is feedback on market design.

The buyer designs the field

Once participation is treated as a sequence, apparently technical choices begin to look like market design. The size of the lot, the timing of proof, and the evidence accepted by the buyer determine which kinds of firms can appear before price or quality is ever compared.

It is tempting to frame procurement as a contest between bureaucratic buyers and excluded suppliers. The reality is less theatrical. Contracting officials also operate under constraints: rules designed to protect fairness, audit requirements, thin staffing, fragmented demand, risk aversion, and pressure to deliver on time. A buyer who cannot easily assess a new supplier may rely on familiar evidence because failure is more visible than missed competition.

The OECD's September work on collaborative procurement in Slovenia is useful here. Centralized and joint approaches can reduce duplicated work and improve capability, but coordination has costs and can reduce flexibility. Aggregating demand may make procurement more professional while making contracts too large for local firms. Efficiency for the buyer can become exclusion for the supplier.

This is a design tension, not a morality play. Unbundling every contract can raise transaction costs. Bundling everything can destroy contestability. The relevant question is proportionality: has the buyer made the package only as large and complex as the outcome requires?

Better supplier access therefore needs buyer-side intelligence too. Which requirements predict performance? Which merely repeat convention? Where could self-declarations replace documents until a later stage? Which lots can stand alone? What does the supplier market actually look like before specifications freeze it out?

Administrative simplification is often dismissed as tidying. Yet a small change in when evidence is required can alter participation materially.

The OECD has highlighted Spain's use of a self-declaration concerning the legal, social, and fiscal status of firms in smaller contracts. Instead of making every bidder produce the complete documentary burden at the outset, the process can defer verification to the stage where it is necessary.

The principle is powerful because it recognizes expected cost. If twenty firms each spend six hours proving facts that only the winner needs fully verified, the system has purchased assurance by imposing 120 hours of supplier effort. Some assurance is essential. But the timing and proportionality of proof determine who can afford to enter.

This is a broader lesson for digital government. Moving a form online does not simplify it if the supplier must still interpret the same rules, locate the same evidence, and repeat the same facts. Digitization changes the medium; simplification changes the work.

A large contract looks efficient because the buyer conducts one competition, manages one supplier, and processes one invoice stream. It also makes assumptions about what kind of organization should deliver the work.

When a city combines landscaping across forty sites, it is not simply buying more landscaping. It is selecting for scheduling systems, vehicles, supervisors, working capital, insurance, and the ability to coordinate a distributed contract. Those may be legitimate performance needs. The package may also exceed what is necessary to achieve them.

The OECD's survey of procurement practitioners found contract size among the most frequently identified SME barriers. Division into lots is the standard response, but lotting is not a free lunch. The buyer must design interfaces among suppliers, evaluate more offers, administer more contracts, and manage inconsistent performance. A small local package may also forgo economies in equipment, routing, and purchasing.

The correct question is not “Can this contract be smaller?” It is “Which integration work genuinely belongs inside one supplier, and which has been bundled because the buyer lacks capacity to coordinate?”

This reframes consolidation. Sometimes a prime contractor creates valuable integration and deserves its margin. Sometimes the prime merely holds the relationship while small subcontractors perform the work under less favorable terms. Award data records small-business participation somewhere in the chain but can obscure where authority, cash, and learning accumulate.

Joint bidding and supplier consortia offer another route. They let firms assemble capacity without surrendering the whole customer relationship to a large prime. Yet collaboration itself requires governance: allocation of scope, liability, pricing, quality, data, and dispute. A recommendation to “team” is not a solution unless the market supplies a way to find compatible partners and form a credible agreement before the deadline.

Contract architecture is thus industrial policy in miniature. It shapes which firms develop which capabilities, which relationships remain direct, and where value concentrates.

Procurement rules aim to constrain favoritism, corruption, and arbitrary judgment. They demand evidence because public buyers owe the public more than intuition. But no requirement is neutral merely because every bidder sees the same words.

A requirement for three comparable government contracts privileges firms that have already received three chances. A high insurance limit favors firms with existing volume or capital. A long financial history screens out young firms regardless of current capability. A narrowly defined certification can exclude equivalent expertise. A short response window favors teams with proposal staff or prior knowledge.

Uniformity is not the same as proportionality. The buyer's duty is to connect each burden to the risk it controls. What failure is this requirement intended to prevent? How predictive is the evidence? Is there an equivalent way to demonstrate the same capability? Can verification occur later, when fewer firms must bear its cost?

This logic does not imply lowering standards. It often produces better standards. A generic demand for years in business may be a weak predictor of performance. A small paid pilot, a work sample, or verified reference may reveal more. Financial capacity matters, but it can be assessed against the cash needs of the actual contract rather than an arbitrary revenue multiple.

The challenge is institutional courage. Familiar proxies protect the buyer from blame. If a supplier with ten years of experience fails, the process appears defensible. If an unfamiliar firm selected through alternative evidence fails, the innovation becomes the story. Rational contracting officers may therefore choose procedural safety over market learning.

Access reform must protect the buyer's ability to exercise and document proportional judgment. Otherwise rules designed to prevent arbitrary exclusion will produce systematic exclusion through rigid proxies.

Trust, portals, and the price of waiting

Formal access still does not make a pathway usable. Firms also need enough trust to invest in pursuit, enough clarity to understand the process, and enough liquidity to survive performance after an award.

Small firms often hear that they need relationships to win contracts. The advice is both true and uncomfortable. Public procurement is supposed to reward fair competition, not insider familiarity. Why should knowing the buyer matter?

Because relationship has at least two meanings. One is improper influence. The other is accumulated understanding. A supplier meeting can reveal how an agency experiences a problem, what implementation failure looks like, which contract vehicle is likely, and how the buyer interprets capacity. A buyer learns whether the supplier listens, follows through, and can explain its limits. None of this should override the published criteria. But it shapes whether a future notice is understood—and whether the market's capabilities shape the notice before publication.

The answer is not to eliminate relationships. It is to democratize legitimate relationship-building: open forecasts, industry days, office hours, transparent question periods, supplier-development programs, and records that help new entrants understand the route. A market without relationships is not necessarily fairer; it may simply preserve hidden relationships for incumbents.

Technology can help by making context portable and explainable. It can show why an opportunity fits, which requirement remains uncertain, what prior decision informed the recommendation, and what a conversation with the buyer needs to resolve. It should never promise influence or imply that scoring can be reduced to a secret formula.

Every opportunity platform is pulled toward volume. More listings make the database look comprehensive. More alerts create engagement. More saved searches create evidence of use. But for a resource-constrained firm, recall without precision is a tax.

Suppose an owner receives 60 weekly alerts and opens twelve. Four are geographically impossible, three contain a hidden certification requirement, two are renewals likely to favor an incumbent, two are plausible but too large, and one deserves serious attention. The platform can report high activity. The owner has lost an afternoon.

The better unit is not the listing. It is the reasoned opportunity: an external notice joined to an internal account of capability, strategy, timing, cash, and relationship. The system should be able to say, “This resembles work you have done, but the insurance requirement and payment timing make it a poor fit unless a partner absorbs those constraints.”

Such a recommendation is harder to build because it requires knowledge of the firm, not merely the market. It also creates a duty of humility. Eligibility rules change. Data is incomplete. A ranking must reveal uncertainty rather than bury it under a score.

A buyer can select a small supplier and still design a contract the supplier cannot finance.

The firm pays employees and vendors while work is performed. It may wait for inspection, acceptance, invoice approval, and a scheduled payment run. The nominal term—30 days, perhaps—begins only after the invoice is accepted. A missing field or disputed deliverable can restart the clock. Large suppliers borrow against a broad balance sheet. The small firm borrows against the owner's credit, delays its own obligations, or declines the market.

World Bank consultations repeatedly identify payment delay as a central difficulty for smaller suppliers. In some surveyed European markets, it was the most frequently reported challenge of government contracting. The problem is not only late payment in violation of a contract. It is the entire cash conversion cycle created by mobilization, retention, acceptance, and administrative verification.

This makes payment design a competition measure. Advances, milestone billing, direct payment to subcontractors, prompt acceptance, invoice-status transparency, and interest on delay can change who can responsibly bid. Faster payment may cost the buyer less than outreach designed to recruit firms that later discover they cannot carry the terms.

Consider a $200,000 service contract with $45,000 in monthly labor. If the first payment arrives seventy-five days after work begins, the supplier may finance more than $100,000 before receiving a dollar. A 10% profit margin does not solve a six-figure timing exposure. The contract is profitable in a spreadsheet and inaccessible in cash.

The supplier may use high-cost online finance to bridge the gap, effectively transferring part of the public contract's value to a lender. The buyer sees a competitive award. The economy sees an avoidable financing chain.

Procurement as a developmental market

These barriers suggest a more ambitious purpose for procurement. A well-designed market does not merely distribute existing opportunity among already legible suppliers; it allows capable firms to become legible through proportionate participation and accumulated performance.

A more contestable procurement market would have several properties. Opportunities would be discoverable early enough for suppliers to prepare. Requirements would be proportional and expressed in language the market can interpret. Evidence already held by government would not be repeatedly requested. Large needs would be divided where division preserves value. Teaming pathways would be visible. Payment terms would reflect the working-capital realities of smaller suppliers. Buyers would learn from who does not bid, not only from who wins.

For small firms, support would shift from generic training toward decision-specific assistance. Instead of another seminar on “how to do business with government,” an owner would receive help answering the next consequential question: Is this worth pursuing, what proof is missing, who could complement us, and what does winning require operationally?

The market opportunity is therefore not another portal. It is an intelligence layer that joins external demand to the firm's actual readiness and learns from pursue, decline, and outcome decisions. Its first promise should be fewer wasted pursuits, not more leads.

Public contracts do more than transfer revenue. They can force a firm to develop cost accounting, quality systems, documentation, cybersecurity, workforce practices, and the ability to deliver at scale. These capabilities can open private markets. Procurement can function as a school.

It can also function as an exam for which no instruction exists. Firms are told to become ready, but readiness is defined by the documents of firms that have already participated. Generic workshops explain acronyms and registration. The actual learning happens while pursuing a real category, usually at the supplier's expense.

A developmental market would create a ladder. Small purchases provide a low-risk first performance record. Buyer feedback explains why a bid was not competitive. Supplier-development support is tied to forecast demand. Larger opportunities become available as evidence accumulates. The firm does not receive a guarantee; it receives a route by which capability can become recognizable.

This is particularly important for firms whose owners have had less access to incumbent networks. Certification can identify the firm and create policy visibility. It cannot substitute for category knowledge, working capital, buyer understanding, or a first credible reference. Counting certified firms without tracing their movement through the funnel risks confusing enrollment with access.

The buyer learns too. A new supplier can introduce local knowledge, innovation, resilience, or a more responsive service model. But learning requires repeated interaction and useful debriefs, not a yearly matchmaking event followed by silence.

The public value of procurement is therefore larger than the purchased deliverable. A well-designed market can create capable suppliers. A badly designed one can consume their capacity before competition begins.

Award dollars will remain essential. They show where public demand ultimately lands. But a serious inclusion strategy would publish a fuller account: number and concentration of participating firms; first-time bidders and winners; median bid cost for representative categories; payment performance; subcontractor terms; reasons for nonparticipation; and the durability of suppliers after entry.

The aim is not to maximize the number of bids. Too many poorly matched bids waste buyer capacity. The aim is a credible, contestable market in which capable firms can recognize a route, assess it at reasonable cost, and compete on evidence related to delivery.

This also disciplines technology ventures. A procurement-intelligence product should be able to show that it improves the funnel, not merely increases alerts. Did firms spend less time on poor-fit pursuits? Did more new suppliers submit compliant bids? Did buyers receive stronger competition? Did payment and performance sustain the relationship? Did the system reveal requirements that unnecessarily narrowed the market?

If it cannot answer those questions, it may be a more elegant portal built atop the same closed market.

Procurement data should therefore return to design. If suitable firms repeatedly stop at the same insurance threshold, payment clause, or bundle, the pattern is not merely supplier failure. It is evidence about the market the buyer has constructed. The learning loop closes only when withdrawals can change future requirements.

Competition is an outcome

The result is a harder standard than formal openness and a more useful one. Competition should be judged by the quality of the pathway that precedes the bid and the durability of the market that remains after the award.

Formal openness is easy to publish. Practical openness must be designed.

A buyer can post a notice, meet a small-business target, and still draw repeatedly from the same supplier set. A firm can be qualified in the ordinary sense and still lack the evidence, liquidity, relationships, or time needed to be recognized as qualified by the process. Neither fact is visible in the award announcement.

The uncomfortable question is not whether small firms can compete. Many plainly can. It is whether the market asks them to become miniature large firms before their merits can be considered.

Access before competition means reducing the avoidable fixed cost of entry while preserving the assurance public buyers owe the public. That is not preferential treatment. It is the work required to make competition real.

Research lineage

Observed evidence. Award totals are large, yet SME participation remains lower than that of larger firms. Contract size, information quality, administrative burden, and qualification requirements are persistent barriers. Collaborative procurement improves buyer capacity but can reduce flexibility.

Interpretation. Opportunity portals solve discovery more readily than fit. The decisive exclusion often occurs in the private bid/no-bid decision before a buyer sees the firm.

Hypothesis. Transparent fit analysis, proportional evidence requirements, and visible teaming routes can widen practical access while saving both supplier and buyer capacity.

Questions carried forward. What makes a recommendation trustworthy? Can a system learn from declined opportunities? Which buyer requirements genuinely predict delivery?

Sources and further reading

1. U.S. Small Business Administration, “Record-Breaking $183B in Federal Contracts to Small Businesses”, January 10, 2025.
2. World Bank, “Firm-level surveys can inform reform efforts to promote more SME participation in public procurement”, January 25, 2024.
3. OECD, SMEs in Public Procurement, 2018.
4. OECD, Enhancing Public Procurement through Collaboration in Slovenia, September 19, 2025.
5. OECD, Implementing the OECD Recommendation on Public Procurement, June 30, 2025.
6. U.S. Census Bureau, Business Formation Statistics, September 2025 release.

Evidence cutoff: September 30, 2025. The facilities-maintenance example is an illustrative composite, not a reported individual case.