How do we size and manage Marketing Development Funds without them becoming partner slush?

What Marketing Development Funds actually are, and why RevOps should own the definition

Marketing Development Funds are dollars a vendor allocates to channel partners to co-fund demand-generation activity — field events, digital campaigns, content syndication, webinars, SDR plays, enablement pushes — that grows the vendor's pipeline through the partner's brand and the partner's audience. The money leaves the vendor's marketing budget, lands in a partner's campaign, and is supposed to come back as registered opportunities. That is the entire theory. Everything that goes wrong with MDF goes wrong because one link in that chain was never actually built.

The single most damaging misconception is that MDF is a reward — something a partner is owed for reaching a tier. The moment funds are framed as an entitlement, they stop being marketing capital and start behaving like a discount routed through the wrong line on the P&L. A partner who believes the money is theirs will spend it the way people spend found money: late, generically, and on whatever is easiest to invoice.

It helps to be precise about where MDF sits relative to the other channel-economics levers, because vendors routinely blur them and then wonder why the program behaves like a price cut:

• Margin / resale discount. The spread a reseller keeps on a transaction. Unconditional, baked into price, spendable on anything. It funds the partner's existence, not the vendor's pipeline.
• Rebate / back-end incentive. A performance payment for hitting volume or strategic targets, usually quarterly. It rewards a result that already happened. It does not fund future demand.
• Co-op funds. Accrued as a fixed percentage of partner purchases and reimbursed against marketing receipts. Backward-looking, entitlement-flavored, and the closest cousin to MDF.
• MDF proper. Forward-looking, discretionary capital attached to a specific approved plan and a specific pipeline target. Of the four, it is the only genuine demand-generation investment.

When a vendor lets MDF drift toward margin behavior — paid regardless of plan or result — it has not been generous. It has quietly converted a marketing budget into a price cut while continuing to carry it as marketing spend. That is the accounting heart of the problem, and it is why this belongs to RevOps rather than sitting in a channel manager's spreadsheet. RevOps owns the definitions that make the number defensible: what counts as partner-sourced, what counts as influenced, how attribution windows are set, and how the resulting ROI line gets reported to finance without being re-litigated every quarter.

"Slush" is the word both sides use, half-jokingly, for MDF that has detached from outcomes. Its signature is recognizable in the field. Accrued balances nobody can explain. Claims for "sponsorship" with no measurable pipeline attached. A 60–80% spike in claim volume in the final two weeks of every quarter, driven by forfeiture panic rather than demand strategy. Partners describing the money in meetings as "our budget" rather than "co-marketing funds." And the tell that matters most: pull last year's claims, sort them into "had a measurable result" and "did not," and if more than a quarter fall into the second bucket, the program is leaking regardless of how good the intentions were.

None of these patterns require fraud. Fraud exists at the edges, but the overwhelming majority of slush is a design failure — a program that never required the partner to say what the money was for, never checked whether that happened, and never reallocated away from what stopped working. Design failures are fixable. That is the good news buried in the diagnosis.

The four-gate lifecycle that keeps funds from becoming partner slush

The structural fix is a lifecycle in which every dollar passes four sequential gates before it leaves the building. No gate, no money. This is the entire governance model, and each gate closes a specific leak.

Gate One — Eligibility. This decides which partners can touch the funds at all, and at what ceiling. Only partners in good standing qualify: current on certifications, no compliance flags, active within the last two quarters. A dormant partner sitting on an accrued balance is the purest form of slush risk there is. Each tier carries a maximum annual ceiling — a workable shape is up to $80K for strategic partners, $25K for premier, $5K for select, and an accrual floor of roughly $500–$2,000 for registered partners. Above the accrual floor, eligibility should require a co-investment match, typically 25–50%. Counterintuitively, the match should *ease* as tier rises: top partners already invest heavily through headcount, certification, and pipeline commitment, while lower-tier partners have less skin in the game and need the higher match as a filter against frivolous spend. Eligibility also needs re-evaluating quarterly, not annually. A partner can be eligible in January and a risk by July — a lapsed certification, a leadership change, two quarters of dormancy. Freeze unspent discretionary funds pending review rather than leaving them available to be dumped on a low-value claim.

Gate Two — The pre-approved plan. This is the most important gate and the one weak programs skip entirely. No dollar gets committed without an approved plan naming a pipeline target. Keep the form short — friction kills participation — but require five things: the activity and audience with an explicit ICP fit argument; the pipeline target in hard units (net-new contacts, MQLs, SQLs, registered opportunities, or sourced bookings); the budget and co-investment split; the timeline including a claim-by date; and the attribution method, whether that is UTM tagging, a dedicated landing page, a registration export, or a campaign code. Score plans against a rubric rather than instinct — ICP fit, target credibility, attribution method, co-investment, and follow-up plan, each 1–5, with a fundable threshold. The follow-up factor is the one teams forget and the one that quietly destroys returns: an event generating 200 leads returns nothing if no one is assigned to call them. Approve within a 3–5 business-day SLA, because slow approvals are the second-biggest driver of slush; partners abandon planned activity and dump the money into easy claims later.

Gate Three — Execution discipline. Two mechanics keep the middle of the lifecycle honest. First, a claim-by deadline tied to the *activity*, not the fiscal quarter — funds approved for a March event must be claimed within 60 days of that event. This single change severs the link between MDF and quarter-end forfeiture panic. Second, mid-flight check-ins for any plan above roughly $15K: is the event on track, are registrations arriving? Treat that check-in as coaching rather than surveillance. A conversation three weeks before an under-registered partner event is the cheapest possible moment to add a co-promoted email, lend a vendor speaker, or extend the campaign window. Accrued balances should expire on a rolling 12-month basis, with automated notices at 90, 60, and 30 days so expiry lands as a prompt to plan rather than a trap. And when a partner gets a plan approved but never executes, release those committed dollars back to the central pool — never roll them into next quarter's ceiling, which rewards non-execution with a bigger balance.

Gate Four — Proof of performance. The claim gate is where slush is finally killed or finally funded. Pay only against two pieces of evidence: proof of spend (itemized third-party invoices for the approved activity, not internal cost transfers) and proof of performance (the deliverable and the result — attendee list, campaign report with leads, content URLs, registration export, UTM analytics, and the count against the approved target). Scale the evidence standard to claim size so small partners aren't drowned in paperwork: under $2,500 needs an invoice plus an activity report; $2,500–$15,000 adds a lead list and ICP confirmation; $15,000–$50,000 adds deal-registration linkage or opportunity IDs; above $50,000 adds a post-activity ROI review with the channel chief. Audit 20–30% of claims at random. The deterrent value vastly exceeds the fraud caught — partners behave differently when they know the evidence might be checked. Watch the rejection rate as a health metric: near zero means the gate isn't real, above 15–20% means the plan gate is approving weak plans or the evidence rules are unclear.

Two governance details make all four gates hold. Separate plan approval (channel marketing) from claim payment (channel finance) — the same person doing both is the control weakness auditors expect to find. And route every exception through one named approver, logged with a reason and reviewed in aggregate quarterly. Exceptions are legitimate and inevitable; the failure mode is exceptions granted informally by whoever the partner emails. If one partner accounts for a disproportionate share of them, that concentration is itself a slush signal.

Sizing math, cost ranges, and the timelines returns actually run on

The cardinal sizing rule: MDF is a percentage of the revenue the channel produces, not a percentage of how many partners you have. Sizing off partner count is exactly how programs end up with a pool far larger than the channel can productively absorb, and an unabsorbable pool is slush waiting for a quarter to end.

Start by splitting partner-attributed revenue into three buckets, because they justify very different treatment. Partner-sourced revenue — deals the partner originated and registered — earns the highest rate, because MDF directly grows it. Partner-influenced revenue — deals the vendor sourced where the partner was materially involved — earns a medium rate, because the funds support the relationship rather than create the demand. Partner-resold or fulfilled revenue, where the partner is transactional fulfillment only, earns very little beyond an accrual floor.

A defensible top-down formula for a hybrid program:

Annual pool = (partner-sourced bookings × 2.0–3.0%) + (partner-influenced bookings × 0.5–1.0%) + (channel revenue × 0.5% accrual floor)

Worked through: a vendor with $40M partner-sourced bookings, $25M partner-influenced, and $90M total channel revenue lands at $1,000,000 + $187,500 + $450,000 = $1,637,500. That number is a ceiling, not a spend target. The discretionary portion (about $1.19M) only releases against approved plans. If partners submit only $800K of fundable plans, you spend $800K plus the floor and return the rest. A program that spends 100% of its pool every year is almost certainly funding slush.

The rate bands are not arbitrary. Push toward the top of each range when the channel is young and needs demand investment, attribution is strong enough to prove returns, partners reliably co-invest, you're seeding a new product or geography, or competitors are outspending you in-channel. Push toward the bottom when the channel is mature and self-sustaining, attribution is weak, partners treat the money as the whole budget rather than a multiplier, or gross margin leaves no headroom. A vendor with weak attribution should run near 1% and spend the difference fixing attribution — you cannot govern what you cannot trace.

Validate top-down with a bottom-up build. Take the tiered partner list and estimate each segment's realistic annual plan: 8 strategic partners at $80K each is $640K; 22 premier partners at $25K is $550K; 140 select and registered partners at a $2,500 floor is $350K — $1,540,000 total. When top-down ($1.64M) and bottom-up ($1.54M) land within 10–15%, the number is credible. A gap over 25% means either attribution is unreliable or you have far more partners than your demand-gen capacity can absorb, which is a slush warning in its own right.

Two structural guardrails on top of the sizing. Hold a 10–15% central reserve unallocated through Q1, then deploy it in Q2–Q3 against the highest-performing activities from the first portfolio review — that converts a passive buffer into an active doubling-down mechanism. And cap any single partner at 15–20% of the discretionary pool; a partner controlling a fifth of the budget has enormous leverage and weak accountability.

How you *release* the pool matters as much as its size. Annual lump allocation is simple but invites year-end forfeiture panic. Plan-triggered release, where no standing balance exists at all, is the most ROI-pure and the most admin-heavy. For most B2B programs the right default sits between them: quarterly tranche release, where roughly a quarter of a partner's ceiling unlocks each quarter and the next tranche is contingent on the prior one being planned and largely claimed. That eliminates the year-end sprint, creates a natural quarterly planning conversation, and gives the vendor a clean off-ramp — a partner who repeatedly fails to plan simply stops receiving tranches, with no awkward clawback required.

On timelines, set expectations honestly with finance. MDF-funded demand does not convert on the quarter it was spent in. Pipeline-to-spend should show up within one to two quarters and target 3–5x. Closed-won-to-spend lags by another quarter or two and targets 1.5–3x. Claim cycle time should hold at 30 days or less from submission to payment, and plan approval SLA hit rate at 90% or better. Utilization belongs in a 70–90% band — and 100% utilization is a warning, not a win, because it almost always means partners burned the budget to avoid forfeiting it.

The tooling cost sits alongside the pool. Below roughly 30–50 partners, a disciplined spreadsheet and a shared drive can run the program, barely. Above that threshold, manual administration is itself a slush generator: claims get lost, balances drift, audits become archaeology. A PRM platform with a native claims module typically runs in the tens of thousands to low six figures annually for a mid-size program. Frame that to the CFO not as marketing software but as a financial control investment — if a $1.6M pool is leaking a conservative 15%, that's $240K a year, and the tool recovers multiples of its cost by enforcing gates a busy human will otherwise skip. When evaluating platforms, the questions that matter are unglamorous: is proof-of-performance a required upload rather than an optional attachment, does it link to the CRM opportunity object, does it structurally prevent a claim against an activity that never had an approved plan, does it fire expiry notices automatically, and does it expose a partner-facing balance and ROI view? Transparency to the partner is itself an anti-slush control.

One caution on tooling: it enforces process, it does not invent strategy. A platform configured around a slush-prone design will simply enforce slush more efficiently. Fix the four gates first, then buy the system that holds them.

Where teams get it wrong — and the failure modes nobody budgets for

The leaks are remarkably consistent across programs, and they're worth stating as concrete failure stories rather than abstractions, because the abstract version undersells how mundane the leakage is.

Funding activity instead of outcome. A vendor approves $30K for a partner to sponsor a regional industry conference. The partner attends, the logo goes on a banner, the booth is staffed. The claim arrives with an invoice and three event photos. No attendee list, no lead-capture mechanism, no follow-up sequence, zero opportunities in CRM. The money bought *presence*, which feels like marketing and generates nothing measurable. The plan never required a pipeline number, so the partner never built one in. This is the most common single failure in the category and it is entirely a design defect — closed by the plan gate.

Accruing without spend discipline. A pure-accrual program banks 1% of every partner's purchases automatically. A healthy mid-size reseller accumulates $48K over 18 months because their resale volume is strong but their marketing function is two slammed people. With a quarter left before a soft nudge, they spend the entire balance on a brochure reprint and an untracked paid-search burst. The dollars left the building; the demand did not. The program treated accrual as automatic and the spend as the partner's problem — closed by rolling expiry, activity-based deadlines, and quarterly true-ups.

Paying claims without proof. A partner submits $12K for a "telemarketing campaign." The invoice is from a sister entity. There is a call log, but it shows internal account-management calls to existing customers, not outbound prospecting. It gets paid because the reviewer checks only that an invoice exists and matches the approved amount — no cross-reference against CRM, ICP fit, or net-new contact creation. The gate existed on paper and was never enforced — closed by mandatory evidence, separation of duties, and the audit sample.

Beyond the three leaks, teams consistently underweight three costs. Opportunity cost typically runs 1–2x the direct waste: dollars locked in low-return activity could have funded plays that actually work. Behavioral corrosion is the quiet killer — a well-run program trains partners to bring their own budget so vendor funds act as a multiplier, while a slush program trains them that the vendor pays for everything, inflating effective CAC and weakening the partner's own commitment. And audit exposure deserves far more weight than finance usually gives it. Under modern revenue-recognition standards, consideration paid to a customer — and a reselling partner can be a customer — may need to be treated as a reduction of revenue rather than a marketing expense unless it buys a distinct service at fair value. Funds that are really a disguised discount can therefore be misclassified, overstating both revenue and marketing OpEx. Automatic accrual on partner purchases also invites the reasonable question of whether quarter-end partner buying is being pulled forward by incentive dollars. And in multi-jurisdiction programs, money flowing without documented business purpose attracts scrutiny under anti-bribery regimes. Proof of performance is also proof of legitimate business purpose. The same governance that kills slush makes the program audit-clean — which is why the CFO should care about this design, not just the channel chief.

Then there's the mirror-image error: over-correcting. If the plan form runs long, the SLA drags, and evidence requirements are punishing, partners simply stop participating. Utilization drops to 40%, which looks like discipline on a dashboard and is actually a different kind of failure — the demand generation the funds existed to create never happened. An aggressive audit posture carries a trust tax too: the best partners, the ones bringing real co-investment and real pipeline, resent being policed like suspects. Calibrate to credible deterrence — random sampling at 20–30%, escalating scrutiny only where there's a history of weak claims — not maximum suspicion.

There's also a diagnostic error worth naming: blaming MDF for downstream failures. A program with strong pipeline-to-spend but weak closed-won has done its job. It generated qualified demand, and the problem has moved to the partner's sales motion — an enablement and coverage problem. Cutting the budget there starves a working demand engine to fix something it cannot touch. Read the metric combinations properly. High utilization with low pipeline-to-spend means slush, and you fix it at the plan and proof gates. Low utilization with high pipeline-to-spend means under-investment, and you fix it with coaching and pre-built campaigns. A fast claim cycle with a rising rejection rate means the plan gate is waving weak plans through. Strong sourced returns with weak influenced returns means influence claims are overstated and the attribution model needs discounting.

Finally, hold metric definitions constant across periods. The fastest way to lose finance's trust is quietly redefining "partner-sourced" or shifting the attribution window between quarters so the ROI line improves. Lock the definitions, document them, report against them consistently, and separate clean sourced returns from directional influenced returns in every board-facing view.

Decision framework: choosing a model, and when MDF is the wrong instrument

There is no universal percentage and no universal model. The right structure depends on channel maturity, attribution capability, partner shape, and whether you're steering a strategic bet or maintaining a mature motion.

Three models cover almost every program. Revenue-share accrual banks a fixed percentage of partner purchases automatically, typically 0.5–1.5% of channel revenue. It's simple and partner-friendly — partners always know roughly what they have — and it's the format most prone to balance-building slush, so it must be layered with expiry and a plan requirement. Discretionary or proposal-based allocation runs 1–3% of partner-sourced bookings and is the most defensible on returns, because every dollar starts life attached to a plan; the costs are administrative weight and a perception of arbitrariness if approval criteria are opaque. Performance-unlock hybrids blend a small accrual floor with a discretionary pool that opens as tier and results warrant, usually landing around 1–2% blended. The hybrid aligns incentives best because partners can see that doing more, and proving it, literally unlocks more budget.

The most-ignored variable in that choice is attribution maturity. If you cannot reliably trace a marketing activity to a registered opportunity, a discretionary program collapses into argument — partners claim influence you cannot disprove, and the channel team rejects claims it cannot evaluate. In that environment, simple accrual with expiry is at least honest about its own limitations. As attribution matures — dedicated landing pages, deal-registration linkage, multi-touch models in CRM — shift weight toward discretionary, because now the proof gate has teeth. Treat the model mix as something that evolves with your data instead of a one-time decision.

Partner shape is the second axis. A handful of high-capability partners with real marketing teams can absorb large discretionary plans and will co-invest. A long tail of small resellers cannot; they have no marketing function and will never write a plan. Forcing everyone into one model is the mistake. Run pre-built, pre-approved campaigns-in-a-box for the tail so their floor dollars still produce tracked demand, and reserve full discretionary planning for partners with capacity to use it well. Product mix is the third axis: a vendor selling one well-understood product through resellers can run leaner and more accrual-weighted, while a vendor pushing several lines or seeding a new one needs the discretionary lever specifically to steer attention toward the strategic bet. It's one of the few tools available to redirect partner focus without renegotiating margin, and a pure-accrual design throws that lever away.

There is a legitimate case for pure accrual, and it deserves stating plainly against the general argument for discipline. In very high-volume, low-touch resale channels — thousands of small partners each transacting modestly — the cost of administering plans and verifying proof genuinely exceeds the slush it would prevent. A simple, well-communicated accrual with rolling expiry beats a discretionary program nobody has the headcount to govern. Match the model to the channel, not to an ideal.

And sometimes the right answer is that MDF should be smaller or absent. If partners are primarily fulfillment and generate little sourced pipeline, the funds will mostly produce slush no matter how well governed; convert the budget to margin or rebate — cleaner and cheaper to run — or redeploy it into direct marketing the vendor controls. For partners who co-sell rather than co-market, the binding constraint is usually AE capacity, not demand-gen dollars. MDF solves a demand-generation problem. If the real problem is sales capacity or enablement, it's the wrong instrument and will look like slush precisely because it's being asked to do a job it was never designed for.

For a team rebuilding from a slush state, sequence the work across a quarter. Weeks 1–3: diagnose — pull two years of claims, classify each as "had a measurable result" or "did not," and let the resulting percentage become the baseline. This phase is non-negotiable and the most commonly skipped; you cannot fix leakage you have not measured. Weeks 3–5: re-size top-down and bottom-up, and approve the pool plus reserve. Weeks 4–8: redesign the four gates, tier ceilings, and controls, then publish the rules openly in the partner portal — surprises breed resentment and gaming. Weeks 6–10: configure the claims workflow or harden the spreadsheet. Weeks 9–12: launch, communicate, ship campaigns-in-a-box, and get the first plans approved. Then run the quarterly portfolio review permanently: defund the bottom quartile — activities returning under 1.5x pipeline-to-spend — unless there's a named, time-boxed strategic exception approved by the channel chief. Concentrating dollars on what demonstrably works, and starving what quietly stopped working, is the single most powerful lever in the entire program.

One last test to run annually: does a dollar routed through a partner generate more pipeline than a dollar the vendor's own marketing team spends directly? If channel-routed dollars consistently underperform, the program needs a reason to exist — usually reach and trust in segments the vendor cannot address directly, whether that's geography, vertical depth, or the SMB long tail. That's a legitimate answer. But ask the question every year, because funds that can neither beat nor credibly complement direct marketing are slush at the program level, no matter how clean the individual claims look.

Related questions

How is MDF different from a partner rebate?

A rebate pays backward for results already achieved and the partner spends it freely. MDF pays forward for a specific approved campaign and requires proof the campaign happened and produced measurable demand. Rebates are cheaper to administer; only MDF actually steers partner marketing behavior.

What co-investment match should we require?

Typically 25–50% above the accrual floor, easing as tier rises — 25% for strategic partners, 50% for select. Partners do not waste their own money, which makes the match the cheapest anti-slush mechanism available. Registered partners on floor dollars alone usually carry no match requirement.

How many claims should we audit?

Random spot-checks on 20–30% of claims, weighted toward larger bands: roughly 30% at $2,500–$15,000, 60% above $15,000, and 100% above $50,000. The value is deterrence rather than fraud caught — partners submit better evidence when they know a check is possible.

Should unspent funds roll over?

No. Accrued balances should expire on a rolling 12-month basis, and approved-but-unexecuted funds should return to the central pool rather than inflating next quarter's ceiling. Rolling forward rewards non-execution with a larger balance, which is precisely backwards.

What utilization rate signals a healthy program?

70–90%. Full utilization almost always means partners burned budget to avoid forfeiture, and very low utilization means the gates have become friction that pushed partners out entirely. Return the unspent remainder to the marketing budget rather than distributing it.

FAQ

What percentage of revenue should we allocate to MDF?

For discretionary programs, 2–3% of partner-sourced bookings plus 0.5–1% of partner-influenced bookings. For accrual programs, 0.5–1.5% of channel revenue. Push toward the top of each band when the channel is young, attribution is strong, and partners co-invest; toward the bottom when attribution is weak or margin is thin. Always validate the top-down number against a bottom-up build of realistic per-partner plans, and treat the result as a ceiling rather than a spend target.

Can we run MDF on a spreadsheet?

Below roughly 30–50 partners, yes — with real discipline about plan records, balance tracking, and claim evidence. Above that, manual administration becomes a slush source in itself: claims get lost, balances drift, and audit trails have to be reconstructed after the fact. A PRM with a claims module that makes evidence a required field and links claims to CRM opportunities pays for itself against even conservative leakage estimates.

How do we handle a partner who never submits plans but has a large balance?

Freeze the balance and open a planning conversation. This is the classic slush scenario, and it usually reflects capacity rather than bad faith — the partner has no marketing function. Offer campaigns-in-a-box so the floor dollars produce tracked demand with minimal effort on their side, and if nothing gets planned across two consecutive quarters, let the rolling expiry do its work with notices at 90, 60, and 30 days.

What evidence counts as proof of performance?

It varies by activity type. Event sponsorship needs an invoice plus an attendee or lead-scan list — "branding only" is the most common rejection reason. Digital campaigns need a campaign report and UTM analytics. Content syndication needs the lead list and content URL, with leads confirmed against ICP. SDR plays need a call or meeting log plus the CRM opportunity IDs created. In every case the requirement is the result, not just the receipt.

Who should approve claims?

Not the person who approved the plan. Separate plan approval (channel marketing, who coaches and judges plan quality) from claim payment (channel finance, who verifies evidence and runs audits). A single approver on both ends is the control weakness external auditors expect to find, and it is the easiest place for weak claims to slide through unchallenged.

How long before we should expect measurable return?

Pipeline-to-spend should reach 3–5x within one to two quarters. Closed-won-to-spend lags by another quarter or two and targets 1.5–3x. If pipeline is healthy but closed-won is not, the gap is in partner sales execution rather than fund design — fix that with enablement rather than by cutting the budget on a demand engine that is working.

Sources

• https://www.forrester.com/blogs/category/channel-marketing/
• https://www.gartner.com/en/sales/topics/channel-partner-management
• https://hbr.org/2011/12/managing-channels-of-distribution
• https://www.fasb.org/page/PageContent?pageId=/reference-library/superseded-standards/summary-of-statement-no-606.html
• https://www.sec.gov/enforcement-litigation/foreign-corrupt-practices-act
• https://www.justice.gov/criminal/criminal-fraud/foreign-corrupt-practices-act
• https://learn.microsoft.com/en-us/partner-center/
• https://partners.salesforce.com/
• https://www.ifrs.org/issued-standards/list-of-standards/ifrs-15-revenue-from-contracts-with-customers/

Related on PULSE

• [How do we build a tiered partner program that rewards scale without collapsing margin?](/knowledge.html?q=q429)
• [How do we run co-sell motions without bottlenecking at account executive capacity?](/knowledge.html?q=q431)
• [What partner enablement curriculum gets partners selling within 30 days?](/knowledge.html?q=q432)
• [What belongs on a board-ready unit economics dashboard?](/knowledge.html?q=q424)
• [How do we calculate and interpret the SaaS Magic Number?](/knowledge.html?q=q418)