Tag: B2B software selection

  • The Complete Guide to SaaS Solution Evaluation

    SaaS solution evaluation is the process a buying team uses to compare software options against a defined use case and reach a decision everyone can stand behind. For cross-functional teams, the hard part isn’t comparing features. It’s getting RevOps, sales, finance, security, legal, and customer success to agree on what matters before the vendors start presenting.

    Most evaluation guides assume one buyer. Real B2B software purchases involve several people with different priorities, and the evaluation often stalls or ends in regret because those priorities were never reconciled. This guide shows how to organize the group, turn disagreement into weights and gates, score fairly, and document the decision.

    Why alignment decides the outcome

    Gartner puts the typical complex B2B purchase at six to ten stakeholders, and Forrester’s broader counts run higher. More people means more priorities. A 2025 Gartner survey found that 74% of B2B buying teams showed what it calls unhealthy conflict, which it defines as conflicting objectives, disagreement over the best course of action, or being overruled by outside decision-makers.

    An earlier Gartner study of 1,120 enterprise technology purchases points the same way. Among buyers who reported high regret, 89% said team members had different, often conflicting objectives for the purchase, compared with 9% of those with no regret. High-regret buyers also took seven to ten months longer on average. These come from large enterprise purchases, so treat the figures as directional for a growth-stage team, but the lesson travels: unresolved conflict at the start becomes regret at the end.

    Set roles and decision rights first

    Before you speak to a vendor, decide who does what. Name one decision owner, usually the RevOps lead or the executive who owns the workflow, who is accountable for the recommendation. Name an executive sponsor to break ties. Then give each function a defined lane. Here’s an illustrative RACI, where R is responsible, A is accountable, C is consulted, and I is informed.

    Activity Sponsor Decision owner Daily users Finance Security and IT Legal and procurement
    Define use case and success metric C A/R C C I I
    Set gates (must-haves) I A C R R C
    Weight criteria I A R R R C
    Score vendors I A R R R C
    Security and data review I C I I A/R C
    Contract and terms I C I C C A/R
    Final decision C A/R C C C C

    Each function should own the criteria it knows best. Security scores security, finance scores total cost, and daily users score usability. Nobody should score everything, and nobody should be asked to rate what they can’t judge.

    Step 1: Agree on one use case and one value scenario

    Write the use case in a card: the metric, the current value, the target, the owner, and the data involved. Our guide to B2B SaaS use case mapping explains the format, and evaluating by revenue bottleneck helps you choose which problem deserves a purchase. Then add a one-sentence group value scenario that everyone can sign: “We’ll know this worked when median first response drops from 19 hours to under one hour, without adding admin headcount.” A shared scenario keeps individual wish lists from taking over. Gartner’s research makes a related point from the seller side: framing for the group’s shared goal improves consensus, while framing for each individual’s agenda damages it.

    Step 2: Gather requirements once, in one place

    Interview each stakeholder separately for 20 to 30 minutes before any group meeting. Ask what a good outcome looks like, what would make them veto a vendor, and what they’ve seen go wrong before. Record answers in one shared requirements document and freeze it at a set date. A requirements document on its fifth version eight weeks in is not diligence. It’s a sign that someone keeps moving the goalposts. Allow changes only through the decision owner, with a reason.

    Step 3: Weight criteria with a priority-points exercise

    Instead of debating weights in a meeting, have each stakeholder allocate 100 points across the criteria privately. Average the results, and look hard at where the spread is widest. Here’s an illustrative example.

    Criterion RevOps Sales lead Finance Security CS lead Average Range
    Workflow fit 30 40 20 15 30 27 15 to 40
    CRM integration 30 10 10 25 15 18 10 to 30
    Security and data 10 5 10 45 5 15 5 to 45
    Total cost 10 15 50 5 10 18 5 to 50
    Adoption and usability 20 30 10 10 40 22 10 to 40
    Total 100 100 100 100 100 100

    The averages give you starting weights. The ranges give you the real conversation. Security and cost show the biggest disagreement, because for the people who own them, they’re not preferences. They’re constraints. The fix is to move them out of the weighted score and into gates: security either passes your review or the vendor is out, and total cost either fits the budget ceiling or it doesn’t. Keep weighting for what’s genuinely tradeable, like fit, integration, and adoption.

    Step 4: Run gates and parallel review lanes

    Security, legal, finance, and operations reviews usually run in parallel, and they often start late. Start them early and run them alongside demos, or they’ll reopen the decision at the end.

    Lane What it checks Typical evidence
    Security and IT Access controls, audit logging, data location, sub-processors, certifications Security questionnaire, architecture overview, compliance reports
    Legal and procurement Data processing terms, liability, termination, renewal pricing caps Draft contract and data processing agreement
    Finance Total cost at 24 months, payment terms, usage-based exposure Written quote at two sizes
    Operations Integration, migration, admin load, support hours Test on your own data

    Step 5: Evaluate on your own data and score independently

    Ask each finalist to run your scenario, using your data and rules, not a polished sample. Label every score by evidence level (claimed, shown, tested, or proven) and cap untested claims. Have each evaluator score independently before any discussion, then discuss only the criteria where scores differ by two points or more. Our guide to SaaS solution scorecards covers the mechanics, including anchored scales and a sensitivity check.

    Step 6: Hold a consensus session

    Once the scores are in, bring the group together for one working session of 60 to 90 minutes. Show the totals, the weights, the evidence levels, and each vendor’s biggest risk. Ask each stakeholder a single question: “Can you live with this recommendation, and if not, what specifically would change your view?” That separates a real objection from a preference. If top scores are within about five points, agree a time-boxed pilot with a success metric, instead of arguing. If someone raises a new criterion late, add it only if it would have changed the weights, and rerun the numbers in front of everyone.

    Step 7: Write the decision down

    A one-page decision memo preserves the reasoning and prevents the same debate at renewal. Include the use case and value scenario, the finalists, scores with evidence levels, the gates each passed, the biggest risk for each, what you’re trading away, the owner, the metric you expect to move, and a review date 90 days out. Add that review to your quarterly RevOps meeting cadence.

    An eight-week plan

    Week Activity Output
    1 Name the decision owner and sponsor; write the use case and value scenario Roles and a shared goal
    2 Stakeholder interviews; freeze requirements; run the priority-points exercise Requirements and weights
    3 Set gates; build a shortlist of two or three Gates and finalists
    4 to 5 Scenario demos on your data; start security, legal, and finance reviews in parallel Evidence and review progress
    6 Sandbox or pilot on the top candidates Tested scores
    7 Independent scoring; consensus session A recommendation
    8 Decision memo; negotiate terms Signed decision and contract

    Treat the timeline as a template. Smaller purchases can run faster, and complex integrations or security reviews may need longer. If your evaluation is for a services partner, not a product, our guide to choosing a HubSpot partner adapts the same ideas.

    Warning signs inside your own buying group

    • The requirements document keeps growing after it was supposed to freeze
    • Reviews get “delegated” back and forth between functions with no conclusion
    • Procurement joins before any vendor is preferred, and negotiates on politics, not price
    • One stakeholder stops responding after a vendor they didn’t prefer advances
    • Different people describe the goal of the purchase in different words
    • The executive sponsor hasn’t been in a meeting since kickoff

    When you see two or more, stop adding demos and fix alignment: revisit the value scenario, the decision rights, and the weights.

    Mistakes to avoid

    • Letting every stakeholder score every criterion
    • Debating weights live, where the loudest voice wins
    • Treating security and budget as preferences instead of gates
    • Starting legal and security reviews after the decision is “made”
    • Skipping the executive sponsor, so ties go unresolved
    • Allowing new criteria late with no rule for changing weights
    • Writing no decision memo, so the reasoning disappears by renewal

    Summary

    SaaS solution evaluation for cross-functional teams is mostly an alignment problem. Gartner finds that most buying teams show unhealthy conflict, and that purchases with conflicting objectives are far more likely to end in regret. Start by naming a decision owner and sponsor, define each function’s lane, and agree on a single use case and a shared value scenario.

    Gather requirements once and freeze them, use a private priority-points exercise to set weights, and convert constraints like security and budget into gates. Run security, legal, and finance reviews in parallel, test on your own data, score independently, and settle close calls with a pilot. Hold a consensus session, write a one-page decision memo, and review the result at 90 days.

    FAQ

    What is SaaS solution evaluation?

    It’s the process a buying team uses to compare software options against a defined use case, using agreed criteria, evidence, and decision rights, so the final choice is explainable and shared.

    Who should be on the evaluation team?

    A decision owner, an executive sponsor, daily users, and representatives from finance, security or IT, and legal or procurement. Keep the core group small, and consult others as needed.

    How do we handle stakeholders who disagree on priorities?

    Collect priorities privately using a points exercise, look at the spread, and turn hard constraints into gates. Discuss the criteria with the widest disagreement, and let the sponsor break any remaining tie.

    Who should make the final decision?

    The person accountable for the outcome, usually the RevOps lead or the executive who owns the workflow, with input from the others. Security and legal should hold authority over their own gates.

    How long should an evaluation take?

    Roughly four to eight weeks for many growth-stage purchases, longer with complex integrations or heavy security review. Starting parallel reviews early is the biggest time saver.

    Should security and cost be weighted criteria?

    Usually they work better as gates, with a pass or fail on your security requirements and a budget ceiling, plus a weighted score for the tradeable differences within those limits.

    What if a stakeholder raises a new requirement late?

    Route it through the decision owner. Add it only if it would change the weights or gates, and if so, rerun the scores openly so everyone sees the effect.

    How do we avoid buyer’s regret?

    Align on one goal early, test on your own data, resolve conflicts before the contract, document the decision, and review results at 90 days. Regret most often traces back to unresolved disagreement at the start.

  • The Complete Guide to B2B SaaS Use Case Mapping

    B2B SaaS use case mapping is the practice of linking each business need to the workflow it affects, the data it depends on, the type of solution that fits, an owner, and a success metric, all in one living document. It’s the missing step between “we have a problem” and “we bought a tool,” and it’s what keeps a growing stack from filling up with software nobody can explain.

    For growth-stage teams, the map does three jobs at once. It shows which problems are worth solving, which kind of solution each one needs (a process fix, a configuration change, automation, AI, or a new platform), and where your current tools already overlap or leave gaps. This guide walks through how to build one.

    Why map use cases before choosing software

    Most software decisions start from a vendor or a category. A use case map starts from the work. That small change has big effects. It exposes duplicate tools doing the same job, makes the choice between automation and AI an explicit decision instead of a marketing label, and gives every evaluation a ready-made set of criteria: the metric, the data, and the owner are already written down.

    It also protects against a common failure: solving the loudest problem instead of the most valuable one. When every use case sits on the same page with the same fields, you can compare them honestly.

    Step one: write a use case card

    Each use case gets one card with the same fields. Here’s a template, filled in with an illustrative example.

    Field What to write Example
    Use case name A short verb phrase Route inbound leads by territory and time zone
    Business need Metric, current value, target Median first response is 19 hours against a 1-hour target
    Workflow affected Where it sits in the lifecycle Lead to first meeting
    Owner One named person RevOps manager
    Data required Sources and fields Lead region, territory table, rep availability
    Current workaround What happens today Manual assignment in Slack
    Solution type From the ladder below Rule-based automation in the CRM
    Success metric and target How you’ll know it worked Median first response under one hour for four weeks
    Constraints Compliance, budget, admin load Must respect consent rules; no new admin hire
    Priority score Calculated in step four 10

    Keep each card to one page. If you can’t fill in the business need with a number, you’ve found a use case that isn’t ready for a purchase yet. It needs investigation first.

    Step two: classify the solution type

    Work down this ladder in order and stop at the first “yes.” The order matters because the cheapest, safest fix comes first.

    Question If yes, the solution type is
    Is the process undefined, or do teams disagree on definitions? Fix the process first: definitions, ownership, required fields
    Can the platform you already own do it with configuration? Native configuration
    Can the rule be written in one sentence, using structured inputs? Rule-based automation
    Does it need judgment on messy or unstructured input? An AI feature, with human review
    Does it span several steps, decisions, and tools? An AI agent, as a supervised pilot
    Does it need data from several systems that disagree? A data unification layer
    Is there a documented gap that none of the above closes? Evaluate a specialist platform

    For the automation-versus-AI rungs, our guide to AI vs. automation in sales explains how to tell the two apart. The short version: if a human wouldn’t need to stop and think, it’s probably automation.

    A worked example: five use cases mapped

    Here’s how the ladder plays out for a made-up growth-stage company. The details are illustrative.

    Use case Solution type Why Success metric
    Route inbound leads across time zones Rule-based automation, native Structured inputs and a rule you can state in a sentence Median first response
    Pass deal context from sales to onboarding Process fix, then required fields and automation Context is missing because nobody agreed what to capture Days from closed-won to first invoice
    Triage prospect replies by intent AI with human review Unstructured text and a judgment call Share of replies triaged within an hour; false flags
    Flag renewal risk from usage Data layer first, then scoring Usage isn’t on the account record yet At-risk accounts flagged before renewal windows
    Improve forecast accuracy Process first, then a forecasting tool if needed Category definitions are inconsistent Forecast error by quarter

    Notice that only one of the five points to buying anything new right away. That’s typical. A map usually shows that most problems need a decision, a definition, or a data fix before they need a vendor.

    Step three: check data readiness

    Every solution type depends on inputs. For each use case, list the data it needs, where it lives, and whether it’s clean and connected. Then give it a readiness rating.

    Readiness What it looks like Next step
    Ready Inputs exist, are consistent, and are connected Proceed to a pilot
    Needs work Inputs exist but are inconsistent or sit in separate systems Clean or connect the data first
    Blocked The inputs aren’t captured at all, such as product usage Instrument the data, or pick a different use case for now

    This step matters most for AI. Predictive tools learn from your history, so inconsistent stages, stale close dates, and duplicate records teach them the wrong lessons. A use case rated “needs work” shouldn’t get an AI tool until it’s rated “ready.”

    Step four: score and prioritize

    Once you have ten or fifteen cards, you need an order. A simple heuristic works: rate each use case from 1 to 5 on impact, readiness, and effort, then calculate impact times readiness, divided by effort. It isn’t science, just a way to force comparison.

    Use case Impact Readiness Effort Score
    Route leads across time zones 5 4 2 10
    Flag renewal risk from usage 4 2 4 2
    Auto-log activity to the CRM 3 5 1 15

    The lowest-impact item ranks first because it’s ready and nearly free. That’s the lesson: quick, well-supported wins build momentum, and low-readiness use cases like renewal risk go on a “prepare” track, where the work is capturing usage data before any tool is chosen. Adjust the scale to fit your team, but write the scores down so the reasoning can be reviewed later.

    Step five: build a coverage matrix

    List your use cases down the side and your current tools across the top. Mark each cell P for primary, S for secondary or overlapping, and M for a manual workaround.

    Use case CRM Conversation tool Onboarding tool Spreadsheet or Slack
    Lead routing P S
    Call insights P
    Onboarding status S P S
    Renewal risk M

    Read the matrix three ways. A row with only a manual mark is a gap, like renewal risk here. A row with several marks is overlap, like onboarding status tracked in three places. A column that covers only one minor use case is a candidate for review at renewal. That last check is how you find shelfware before the invoice arrives again.

    Keep the map alive

    A map that’s built once and filed away goes stale within a quarter. Assign a single owner, usually RevOps, and review it on a schedule. Add new use cases as they surface, retire ones that were solved, and update scores when data readiness changes. Recording the reason for each decision, including “not now,” saves you from relitigating it six months later. If you already run a RevOps meeting cadence, the quarterly review is the natural home for this.

    When a use case reaches the evaluation stage, the card already holds your criteria. For running the evaluation itself, see how SaaS teams evaluate use case solutions by revenue bottleneck, and for choosing between optimizing, extending, replacing, or waiting, our buyer’s guide to B2B SaaS solutions.

    Mistakes to avoid

    • Writing use cases as features (“we need a dashboard”) instead of needs with a metric
    • Skipping the process-fix rung, then automating a disagreement
    • Choosing AI for a task a one-sentence rule could handle
    • Ignoring data readiness until after a tool is purchased
    • Letting the map belong to everyone, so nobody updates it
    • Mapping only new needs and never checking the tools you already own

    Summary

    B2B SaaS use case mapping turns software decisions into a repeatable process. Write a card for each need, with a metric, an owner, the workflow, the data required, and the current workaround. Classify each one on a ladder that starts with process fixes and native configuration and moves through automation, AI, agents, data unification, and specialist platforms only when the earlier rungs don’t fit.

    Rate data readiness, score use cases on impact, readiness, and effort, and build a coverage matrix to expose gaps, overlaps, and shelfware. Keep one owner, review the map every quarter, and record why decisions were made. Most maps reveal that only a few needs call for a new purchase right away, which is the point.

    FAQ

    What is B2B SaaS use case mapping?

    It’s the practice of documenting each business need alongside its workflow, data, solution type, owner, and success metric, so software decisions are tied to specific work instead of vendor categories.

    How many use cases should we map?

    Start with ten to fifteen of the most painful ones. A longer list is harder to maintain and usually includes items that don’t deserve attention yet.

    Who should own the use case map?

    Usually RevOps, with input from sales, marketing, customer success, and finance. Each individual use case should still have its own named owner for the workflow it affects.

    How is a use case map different from a tech stack diagram?

    A stack diagram shows which tools you have and how they connect. A use case map shows which jobs need doing and what solves each one. Overlaying the two, as in the coverage matrix, is what reveals gaps and overlap.

    When should a use case get AI instead of automation?

    When it needs judgment on messy or unstructured input, such as reading a prospect’s reply, and when a person will review the output. If a one-sentence rule on structured data would work, use automation.

    How often should we update the map?

    Quarterly works for most growth-stage teams. Also update it when a major workflow changes, a tool is added or removed, or a data source becomes available that changes readiness.

    What if a use case is high impact but not ready?

    Put it on a prepare track. Do the data and process work first, such as capturing usage on the account record or agreeing on definitions, so a tool has something reliable to work with when you evaluate one.

  • A Buyer’s Guide to B2B SaaS Solutions in 2026

    Somewhere in your stack there’s a tool nobody opens. It was bought with real enthusiasm, launched with a kickoff call, and now costs a few thousand dollars a year while the team quietly goes back to a spreadsheet. Nearly every growth-stage team has at least one.

    The usual verdict is that the tool was bad. More often the decision was thin: the problem was never written down, nobody owned the rollout, and the first real test happened after the contract was signed.

    This guide is for revenue operations leaders at growth-stage SaaS teams who want a steadier way to approach B2B SaaS solutions in 2026. It covers the signals that tell you whether you need a tool at all, a few decision frameworks for choosing between paths and vendors, and what has changed about business software evaluation lately.

    Start by asking whether you need a tool at all

    Buying software feels like progress, which is exactly why it’s a popular answer to problems that aren’t software problems. Before you open a single vendor tab, look for these signals.

    Signal What it usually means Sensible next move
    Several people repeat the same manual task every week A real workflow gap Write the use case and evaluate tools
    Two teams report different numbers for the same metric A definitions problem Agree on definitions first; the tool can wait
    The platform you own could do it, but nobody configured it A setup gap Optimize what you have
    A workflow has outgrown a native feature, such as routing across regions A genuine platform limit Evaluate a specialist
    A new executive wants the tool from their last company Familiarity, not a signal Ask for the use case, then decide
    Reps say they hate the current tool Could be adoption, training, or fit Find out which before replacing anything

    Only the first and fourth rows point toward a purchase. The rest point toward conversations, and conversations are cheaper.

    Four paths, not one

    Most evaluations treat “buy a new tool” as the only option. There are four, and growth-stage SaaS teams tend to underuse two of them.

    Path Fits when Main risk
    Optimize The platform can do the job, but setup or process is lacking Time spent configuring something you’ll soon outgrow
    Extend One workflow has outgrown a native feature Another contract and another integration to maintain
    Replace A platform limit is blocking revenue, or costs no longer scale Migration cost, and carrying old process problems into new software
    Wait The problem is real but small, or the team can’t absorb change right now The problem grows while you wait

    A quick test for the first path: could a capable new hire, given a week and the vendor’s documentation, make what you already own do this job? If yes, optimize.

    The last path gets a bad reputation as procrastination, but waiting with a written trigger is a real decision. “Revisit when we pass 40 reps or open a second region” gives you a reason to look again and a reason to stop worrying until then.

    Scoring vendors without fooling yourself

    Once you’ve decided to evaluate SaaS tools, a weighted scorecard keeps the loudest voice in the room from making the call. Pick five criteria that matter for this particular use case, give them weights that add up to 100, and score each vendor from 1 to 5 using your own data, not the vendor’s demo.

    Criterion Weight Vendor A Vendor B
    Fit to the workflow 30 4 5
    CRM integration depth 25 5 3
    Time to value 15 3 4
    Total cost at 24 months 15 3 4
    Support and adoption 15 4 3
    Weighted total (out of 100) 100 79 78

    To get the totals, multiply each score by its weight, add the results, and divide by five. Vendor A lands on 79 and Vendor B on 78.

    That isn’t a winner. It’s a tie, and a one-point gap on a scorecard built from judgment calls shouldn’t decide anything. Ties are what pilots are for. If the totals are far apart, believe them. If they’re close, run both against one real workflow for a few weeks and let the results break the tie.

    One habit makes this work: set the weights before you see pricing or sit through any demos. Teams that weight criteria afterward often discover the criteria happen to favor whoever impressed them most.

    What’s different about buying in 2026

    A few things have shifted enough to change how you run an evaluation.

    • AI is everywhere, so ask what it actually does. Nearly every vendor now lists AI features. Ask whether they’re included or a paid add-on, whether they work on your data, and whether the tool can explain its output. A score nobody can explain is hard to defend in a forecast meeting.
    • Pricing is moving toward usage and credits. Per-seat pricing is easy to model. Usage pricing is harder, so estimate cost at the volume you expect in two years and ask what happens when you go over.
    • Platforms keep absorbing point tools. The CRM or marketing platform you already own may cover part of what a specialist sells. Check the overlap before you add anything.
    • Your data has to be able to leave. Ask how you’d export everything, in what format, and how quickly. It matters most on the day you want to switch.

    Terms worth negotiating before you sign

    Growth-stage teams often have more leverage than they assume, especially on a first contract. A short list worth raising:

    • A paid pilot, or an opt-out in the first 60 to 90 days
    • A cap on renewal price increases
    • Written implementation scope, including who does what and by when
    • Support hours and response times that match your time zones
    • A clear process and format for exporting your data
    • A reasonable notice window before auto-renewal

    None of these are unusual requests. A vendor who resists all of them is telling you something about how the relationship will go.

    Put the review date in the calendar before you buy

    The purchase isn’t the finish line. Before you sign, write down the one measurable outcome you expect, name the person accountable for it, and book a review 90 days out.

    At the review, ask the question you started with: did the number move? If it didn’t, work out whether the cause is the tool, the setup, or the process, and decide with that in hand rather than discovering it at renewal time.

    Mistakes that keep repeating

    • Treating a vendor demo as the start of the process instead of a written use case
    • Replacing a tool to fix what is really a process problem
    • Adding several tools at once, so nobody can tell which one made a difference
    • Quoting list price and forgetting implementation, admin time, and add-ons
    • Skipping the pilot because the vendor’s references sounded good

    Summary

    Choosing B2B SaaS solutions well starts before any vendor conversation. Look for real signals, a repeated manual workflow or a native feature that has been outgrown, and treat the rest as prompts to fix definitions, setup, or adoption first. Then pick from four paths: optimize, extend, replace, or wait, remembering that optimizing and waiting are underused.

    When you do evaluate vendors, set weighted criteria before seeing any demos, score on your own data, and treat close scores as a tie to be settled by a short pilot. In 2026, ask what AI features really do, model usage-based pricing at future volume, check overlap with your existing platform, and secure data export rights. Negotiate a pilot or opt-out, book a 90-day review, and measure the one outcome you bought the tool to move.

    FAQ

    How many new tools should a growth-stage team add in a year?

    There’s no magic number, but adding one meaningful tool at a time lets you tell what actually worked. Several at once makes the results impossible to untangle.

    Should we choose a suite or best-of-breed tools?

    Start with what your core platform does natively. Add a specialist only where you can point to a documented gap, and make sure it writes its data back to your main system.

    How long should an evaluation take?

    Roughly four to eight weeks from written use case to decision, including a short pilot. Longer than that and momentum usually fades.

    What if the team disagrees on the choice?

    Go back to the scorecard and the weights you agreed on beforehand. If the totals are close, run a pilot instead of arguing about it.

    When is it time to replace a tool instead of extending or optimizing it?

    When it’s blocking revenue or its costs no longer scale, and you’ve confirmed the real problem isn’t setup or process. Otherwise the same issues tend to follow you to the replacement.

    Who should own the final decision?

    Whoever is accountable for the outcome, usually the RevOps lead or the executive who owns the workflow. Daily users, finance, and security should all have input before it’s made.