Guide
Vendor selection process: 9 steps for small and mid-size companies
The vendor selection process is how a company goes from a business need to a signed vendor: define the need and budget, check the offers you already have, build a long list, choose between an RFI, RFP or RFQ, shortlist, run demos and reference calls, score the finalists, negotiate and record the decision, then notify the vendors who lost and onboard the winner. For a small or mid-size company it takes from about a week for a small, easy-to-reverse purchase to three months or more for a strategic one.
Updated November 17, 2026 ยท 15 min read
Key takeaways
- Scale the process to the purchase: a $3,000 tool needs two quotes and an email; a multi-year contract that touches customer data needs every step.
- Write requirements before you talk to any vendor. Without them, the best salesperson defines your need for you.
- Step 2 is the one most teams skip: check the vendors, contracts and offers you already have on file before searching.
- The person who will use the product or service every day must score the finalists. The budget holder approves; they should not pick alone.
- Fix scoring weights before proposals arrive, and write a one-page decision record when you choose.
- Tell the losing vendors. They are your fallback list and next year's bidders.
The vendor selection process at a glance
Most guides to the vendor selection process are written for procurement departments with category managers and a sourcing platform. A 30-person company has none of that: it has a founder, a team lead who needs the thing, someone who pays the invoices and perhaps an outside IT provider. These nine steps keep the parts that protect you and cut the ceremony. Each step produces one written output, so the process leaves a trail even when it takes a week.
| Step | Written output | Shortcut for small purchases |
|---|---|---|
| 1. Define the need and budget | One-page requirements | Five bullet points in an email |
| 2. Check what you already have | List of existing vendors and past offers with prices | Never skip: it takes 15 minutes |
| 3. Build a long list | 5 to 8 screened candidates | 3 candidates |
| 4. Choose RFI, RFP or RFQ | Request document sent with a deadline | A quote request by email |
| 5. Shortlist | 2 to 4 finalists, signed-off scoring weights | 2 finalists |
| 6. Demos, trials and references | Demo notes, trial results, reference call notes | One reference call or a free trial |
| 7. Score and compare | Completed scorecard | Side-by-side quote comparison |
| 8. Negotiate and record the decision | Agreed terms, decision record | Three-line decision note |
| 9. Notify, sign and onboard | Signed contract, notices sent, onboarding started | Short thank-you email to the others |
Step 1: Define the need, the budget and the owner
Write down what problem you are solving before you talk to any vendor. Vendors are good at reframing your need around what they sell; a written requirement is the only defense. Split it into must-haves (pass/fail: miss one and the vendor is out) and nice-to-haves (scored). Add a budget range, a deadline and the names of the people who decide. If you cannot fill in the budget line, stop and get one: a selection without a budget ends with the most expensive vendor and a delayed approval.
Five to eight must-haves is plenty. If you have twenty, most of them are preferences, and a long pass/fail list will eliminate every vendor or tempt you to quietly ignore it.
Step 2: Check what you already have on file
Before searching the market, search your own company. Four places to look:
- Existing vendors. Can a vendor you already pay do this, under a contract you already negotiated? Adding scope to a trusted vendor often beats onboarding a new one, and it is a useful price benchmark even when it does not.
- Existing contracts. Check for exclusivity clauses, minimum commitments or bundled services you are already paying for and not using.
- Past selections. Did another team evaluate this category in the last two years? Their scorecard, shortlist and reasons for rejecting vendors save weeks.
- Offers you received. Proposals, quotes and cold pitches that arrived with a price are free market research. The problem is that they sit in individual inboxes and get deleted.
This is the step PitchBin is built for. Teams forward vendor emails to save@pitchbin.com, and each one is filed in a shared company catalog with the vendor company, service, quoted price and category, plus the team's notes and ratings. When a need appears, searching "payroll" or "penetration test" returns every offer anyone in the company has received, with the price as originally quoted (a sent pitch is an immutable snapshot). That turns step 2 from asking around into a search, and it often hands you half the long list for step 3 with price benchmarks attached.
Step 3: Build a long list of 5 to 8 vendors
Combine sources so you do not end up with only the vendors who advertise hardest:
- Offers and vendors from step 2
- Referrals from peers at companies of your size, ideally people who would tell you about a bad experience
- Industry associations, trade directories and your accountant, lawyer or IT provider, who see many vendors across clients
- Search and review research, read with skepticism
- An open call: publish what you are looking for and let vendors come to you. In PitchBin a company can publish a buying need as its own pitch link, and offers that answer it are tagged and counted, so they arrive structured (company, service, price) instead of as a pile of emails.
Screen the long list quickly before anyone books a demo: is it a real, registered company with a working website and a track record, does it serve companies your size, does it plausibly meet every must-have? Ten minutes per vendor is enough at this stage. Deep checks come in step 6.
Step 4: Choose an RFI, RFP or RFQ (or none)
The request you send decides what you get back. Pick the lightest format that still gives you comparable answers. The full comparison, with templates, is in our RFI vs RFP vs RFQ guide.
| Format | Use it when | What you send | Response time to allow |
|---|---|---|---|
| Simple quote request | Small, standard, easy to switch purchases | An email with the spec, quantity and deadline | 2 to 5 working days |
| RFQ (request for quotation) | You know exactly what you want and price is the main difference | Detailed spec, quantities, delivery terms, a fixed price format | 1 to 2 weeks |
| RFI (request for information) | You do not yet know what solutions exist or what they cost | The problem, 10 to 15 questions, a request for indicative pricing | 1 to 2 weeks |
| RFP (request for proposal) | Several approaches could work and you need vendors to propose one | Requirements, evaluation criteria, timeline, pricing template, contract basics | 2 to 4 weeks |
Always send the same document to every vendor, with a pricing table they must fill in your format. If each vendor prices in its own structure, step 7 turns into a week of untangling which quote includes what.
Step 5: Shortlist 2 to 4 finalists and lock the weights
Drop every vendor that failed a must-have or did not respond by the deadline (a vendor that misses the deadline for its own sales pitch will miss yours). Keep two to four finalists: fewer gives you no leverage, more and the demos take over the team's month.
Before any finalist presents, agree the scoring criteria and weights with the budget holder and write them down. Weights decided after you have seen prices and demos tend to bend toward the vendor someone already likes. Our vendor evaluation criteria guide has 30 criteria with scoring anchors, weights for software, agency and physical supplier purchases, and a downloadable scorecard.
Step 6: Demos, trials and reference calls
Run a scripted demo
Send each finalist the same three to five real scenarios from your work ("show us how a refund request from a customer moves through your tool", "walk us through how you would plan our Q1 campaign") and ask them to demo those, not their standard deck. Ask that the people who will do the work, not only sales, attend. Score each demo the same day while it is fresh.
Trial or pilot where you can
For software, a trial with your own data beats any demo. For agencies and contractors, a small paid pilot (one deliverable, fixed price) tells you more about quality and communication than three meetings. For goods, ask for samples and check them against your spec.
Call references, including one the vendor did not pick
Ask each finalist for two references at companies like yours, and ask whether they will connect you with a former customer. Questions that get past the polite answers:
Run the formal checks in parallel for the finalists you are serious about: company registration, financial health, insurance, security and data protection evidence. The full list, in order of priority, is in our vendor due diligence checklist.
Step 7: Score and compare the finalists
Each evaluator scores alone first, with one line of evidence per score. Then meet, and discuss only the criteria where scores differ by two points or more; that is usually where one person saw something the others missed. Calculate the weighted totals, then check how fragile the result is: if a small change in weights flips the winner, the decision is a judgment call and should be recorded as one.
Before scoring cost, normalize the quotes: same scope, same term, same volume, with setup, training, support tier and renewal increases added in. Compare three-year total cost, not the first invoice. A quote that looks far cheaper often turns out to exclude onboarding or to cap usage below what you need.
Step 8: Negotiate and write the decision record
Tell the preferred vendor they are preferred, not that they have won, and keep the runner-up warm until the contract is signed. Price is one lever among several, and for small companies often not the most valuable one:
- Term and exit: a shorter initial term, termination for convenience, no auto-renewal or a short notice period
- Price protection: a written cap on renewal increases
- Payment terms: monthly or quarterly instead of annual upfront, or your standard invoice terms
- Service levels: response and resolution times with service credits
- People: named key people written into the contract, with your approval needed to replace them
- Data and exit assistance: full export of your data or work product in a standard format at no cost
Then write the decision record. It takes ten minutes and answers every future "why did we pick them?" question, including at renewal, during an audit, or when the person who ran the selection has left.
Step 9: Notify the other vendors, sign and onboard
Once the contract is signed, tell every shortlisted vendor the outcome within a few days. A short, specific note ("you were close on capability; price was the deciding factor") keeps the relationship usable, and the runner-up is your fallback if the winner fails in the first six months. Vendors that were ghosted rarely bid seriously the next time. Templates for each situation are in our vendor rejection letter guide.
Onboarding is part of selection, not an afterthought: collect tax and payment details through a verified channel, sign the data processing agreement where the vendor will handle personal data on your behalf (under the EU and UK GDPR, Article 28 requires a written contract with every processor), grant only the system access the vendor needs, put the renewal notice date in a shared calendar, and set the KPIs and first review date from the decision record. The vendor onboarding checklist covers each item. This guide is practical guidance, not legal advice.
How long vendor selection takes, by purchase size
The process should be proportional to what you lose if you choose badly: the money, the switching cost and the risk to your data or operations. The thresholds below are examples for a company of roughly 10 to 250 people; set your own in your purchasing policy. Elapsed times are typical ranges, mostly spent waiting for vendor responses and calendar slots, not working hours.
| Purchase size | Example threshold (annual spend) | Vendors compared | Request type | Sign-off | Typical elapsed time |
|---|---|---|---|---|---|
| Small, easy to switch | Under $10,000 | 2 to 3 quotes | Quote request by email | Budget holder | 1 to 2 weeks |
| Medium | $10,000 to $50,000 | 3 finalists | RFQ or short RFP | Budget holder and finance | 3 to 6 weeks |
| Large, or touches customer data | $50,000 to $250,000 | 5 to 8 long list, 3 to 4 finalists | RFI, then RFP | Founder or CEO, finance, IT/security | 6 to 12 weeks |
| Strategic | Over $250,000, multi-year, or core to operations | 3 to 5 finalists plus a pilot | RFI, RFP, paid pilot | Leadership team, board where required | 3 to 6 months |
Two rules hold at every size: compare at least two options (even if one is "the vendor we already use"), and write at least three lines of decision record. A purchase that holds personal data or system access moves up one row regardless of price.
Who does what: a RACI table for a small team
RACI marks who is Responsible (does the work), Accountable (owns the outcome, one person per row), Consulted (gives input before) and Informed (told after). In a small company one person often holds two roles; keep the roles distinct anyway, especially the line between the person who uses the product and the person who pays for it.
| Activity | Day-to-day owner (team lead) | Budget holder (founder or dept head) | Finance | IT / security | End users (2 to 3) | Legal (often external) |
|---|---|---|---|---|---|---|
| 1. Requirements and budget | R | A | C | C | C | - |
| 2. Check vendors and offers on file | A, R | I | C | C | - | - |
| 3. Long list and screening | A, R | I | - | C | C | - |
| 4. RFI / RFP / RFQ | A, R | C | C | C | - | - |
| 5. Shortlist and weights sign-off | R | A | C | C | C | - |
| 6. Demos, trials, references, due diligence | A, R | I | R (financial checks) | R (security review) | R (trial) | - |
| 7. Scoring | A, R | I | R (cost) | R (risk) | R | - |
| 8. Negotiation and contract | C | A | R | C | - | R (contract review) |
| 8. Decision record | R | A | I | I | I | - |
| 9. Notify other vendors | A, R | I | - | - | - | - |
| 9. Onboarding | A, R | I | R (vendor setup, payment details) | R (access) | I | C (DPA) |
Note where the budget holder is accountable: requirements, weights, contract and the final decision. Note where they are not: scoring. If the founder wants to overrule the scored result, that is their right, but the reason goes in the decision record.
Common vendor selection failure modes
Most bad vendor choices are not caused by bad vendors. They are caused by a process that let the wrong signal decide. The patterns that come up again and again in small companies:
- Choosing the loudest pitch. The vendor with the most persistent salesperson or the slickest demo wins, because nobody wrote down what "good" meant. Fix: written requirements, a scripted demo with your scenarios, scores backed by evidence.
- No written requirements. Each vendor redefines the need around its product, and the team ends up comparing three different solutions to three different problems. Fix: the step 1 one-pager, approved before the first call.
- The CEO picks without the user. The founder sees one demo, likes the vendor, signs. The team that has to use it every day was never asked, and adoption stalls. Fix: the day-to-day owner and end users score; the budget holder approves and records any override.
- Weights set after the prices are known. The math gets built to justify a choice already made. Fix: sign off weights in step 5.
- Only one quote. "We already know them" means no benchmark and no leverage. Fix: at least two options, even for small purchases.
- Apples-to-oranges quotes. One price includes onboarding and support, another does not. Fix: one pricing table in your format for every vendor, and three-year total cost.
- References that were never called, or only the vendor's favorite customer. Fix: two comparable references and, where possible, a former customer.
- Ignoring exit terms. Auto-renewal with a 90-day notice window and no data export turns a one-year deal into three. Fix: negotiate exit in step 8 and calendar the notice date in step 9.
- An evaluation with no end date. Vendors lose interest, the need changes, the team loses patience and picks whoever is still answering. Fix: put a decision date in the step 1 document.
- Undeclared conflicts of interest. The vendor is a friend's company or a former employer. That is not automatically disqualifying, but it must be declared and that person should not score alone.
- Ghosting the losers and losing the file. Next year the same need returns, nobody remembers who bid or at what price, and the whole process restarts from zero. Fix: notify vendors and keep the shortlist, quotes and scores somewhere the team can search.
Frequently asked questions
What are the steps in the vendor selection process?
Define the need and budget, check the vendors and offers you already have, build a long list, choose an RFI, RFP or RFQ, shortlist two to four finalists, run demos, trials and reference calls, score the finalists, negotiate and write a decision record, then notify the other vendors and onboard the winner.
What are the main vendor selection criteria?
Total cost of ownership, quality, delivery, capability (experience, named team, technical fit), risk and security, financial stability, the working relationship and ESG. Weight them by purchase type: security and integration for software, people and quality for agencies, delivery and defect rates for physical suppliers.
How do you choose a vendor for a small business?
Write five to eight must-haves and a budget, check whether an existing vendor or a past offer already covers it, get at least two comparable quotes, call one reference per finalist, and let the person who will use the product or service score the options. Record the decision in a few lines.
How long should a vendor selection process take?
As a rule of thumb, one to two weeks for small, easy-to-switch purchases, three to six weeks for medium ones, six to twelve weeks for large purchases or anything that touches customer data, and three to six months for strategic, multi-year contracts.
How many vendors should you evaluate?
Start with a long list of five to eight, screen it quickly, and take two to four finalists into demos and scoring. Fewer than two gives you no benchmark; more than four usually costs more team time than it saves.
Is the supplier selection process different from vendor selection?
The steps are the same. Supplier selection for physical goods puts more weight on samples, capacity, lead times, defect rates and supply chain checks; vendor selection for software and services puts more weight on trials, security, data protection and the named team.
Keep reading
Start your next vendor selection with a search, not a blank page.
Forward vendor offers to save@pitchbin.com and your team builds a shared catalog of every offer received, with company, service and quoted price. Free for companies receiving offers.