Tech donors are an increasingly vital source of philanthropic capital. To reach them, what you need is a clear account of what a dollar buys, an honest read on what it changes, and the right giving vehicle. This playbook covers all of them.
For nonprofit development and program teams. General education, not tax advice. Read the disclaimer.
People who build software, hardware and data products for a living tend to read an appeal the way they’d read a design document. They look for the mechanism, the unit, and the point where the claim could turn out to be wrong. And when they give, the money rarely arrives as a single check: it comes as shares, through a donor-advised fund (DAF), doubled by an employer, on a timetable set by vesting dates and brokerage cutoffs.
This playbook works through both halves. Parts one and three cover the case: how these donors think, what they fund, and how to write an ask and a memo that hold up to questions. Parts two and four cover the money: how a gift is assembled, what the 2026 tax changes did to it, and when to ask. Part five is two working tools. The drawings are there to show structure and help readers to understand. Where one is a schematic rather than data, it says so.
Four things fundraising teams actually come here to do. Each links to the section that covers it, so you can take the piece you need and leave the rest for later.
Three patterns we often see in donors who work in software, hardware, product or data. Treat them as working hypotheses to test against your own file, not as established traits of a demographic. Your giving history and the interests your donors have actually shown will beat any generalization on this page.
Tech donors tend to see social problems as systemic inefficiencies, like coordination failures or broken feedback loops, rather than as isolated misfortunes. They’re wary of spending that gets used up, like emergency supplies, which address symptoms but not root causes. They favor work that permanently lowers friction and reduces costs over time.
What this means for your ask. “We need $50,000 to keep the pantry stocked through winter” answers a different question than the one they’re asking. Say what keeps sending people to the pantry, what your program does about that, and what would still be true next winter if the gift lands.
Relief refills the same need each winter. Work on the cause is judged by whether the need shrinks. Both are legitimate; this donor is usually asking about the second.
Aggregate budgets tend to land poorly. What works is a per-unit figure: what one outcome costs you today, what it would cost after the investment, and how you know the difference is real rather than selection effects.
The number that most often goes wrong. Cost per participant isn’t cost per outcome. If 100 people enroll at $2,000 each and 60 finish, your cost per completion is $3,333. And if 20 of those 60 would have finished anyway, your cost per additional completion is higher still. Scroll through the example below to see how much higher, and why leading with it works.
The example program below: 100 enrollments at $2,000 each.
Say a training program costs $200,000 a year and enrolls 100 people. Divide one by the other and you get $2,000 per participant. That’s the number most appeals quote.
Each square is one participant.
Forty people leave before the end. The outcome you’re funded for is completion, so the cost per completion is $200,000 divided by 60: about $3,333.
Suppose 20 of those 60 would have finished without the program, because they were already motivated, already had support, or already had the skills. You didn’t cause those completions.
That leaves 40 additional completions, so the cost per additional completion is $5,000. It’s the largest of the three numbers and the only one that answers the donor’s question.
Section 09 works through the same arithmetic in a full memo, including what a comparison group can and can’t prove.
Many tech professionals define themselves by what they can fix. In our experience, being approached first as a wallet puts them off, and being asked for their expertise on a real problem opens the door. The gift may or may not follow, so treat the conversation as worth having on its own terms.
What to actually ask them for. Not a hackathon. Ask them to pressure-test your cost model, poke holes in your scale plan, look at whether your outcome data actually supports your claim, or tell you what they’d want to see before funding this. Those are 30-minute conversations that a senior engineer or product manager is genuinely good at.
Development teams often conclude that reaching tech donors means proposing an app. The opposite is closer to true. People who build software for a living have sharp instincts against unnecessary software, and an app for a soup kitchen reads to them as a warning sign, not a signal of sophistication. What travels isn’t the technology. It’s the reasoning.
Four examples of non-technical work that has attracted significant technology-sector money. What they share is an unusually clear account of what a unit costs and how it’s verified. Each drawing shows the shape of that reasoning.
GiveDirectly moves money to people in poverty with minimal intermediation and has published randomized controlled trials on the results. The pitch is that fewer steps between donor and recipient means less loss, and that the claim is testable.
Nets are physical objects handed to households. GiveWell estimates roughly $4 to $6 per net, and somewhere between about $3,000 and $8,000 to avert a death in the places it funds, as of its December 2023 analysis. The range is part of the credibility.
Housing supply advocacy attracts tech money because one zoning or permitting change can shape housing production for decades, at a cost far below building units directly. California YIMBY launched in 2017 with about $500,000 from tech executives and backed later rounds of the state’s ADU reforms. Be careful with the arithmetic: name the rule, the place and the plausible range, not a multiplier.
Youth Villages’ LifeSet gives young people leaving foster care a specialist, and Ballmer Group has funded it in Washington. In MDRC’s randomized trial, earnings rose 17% and the chance of homelessness fell 22% after a year, with no change in education or crime. It costs about $8,400 extra per participant; cost-benefit studies disagree on its payoff.
This is foundation-scale money; the lesson for your file is the kind of evidence it rewarded.
A note on measurable outcomes. Measurable outcomes are one good reason to give, not the only one. Immediate relief, dignity, long relationships and the knowledge a community holds about itself are legitimate reasons too, and some of the most important work resists a unit cost entirely. If a number would misrepresent your work, don’t manufacture one. Say plainly what you do and why it matters.
Three sources sit behind most large gifts from tech employees. Each has its own paperwork, timing and failure mode.
A DAF is a charitable account. The donor contributes, takes the deduction at that moment, and recommends grants out of it later. The 2025 Annual DAF Report, covering fiscal 2024, puts DAF assets at $327.87 billion across 3.59 million accounts, with $64.60 billion granted out and a payout rate of 25.2%. The report is now published by the Donor Advised Fund Research Collaborative, which took it over from the National Philanthropic Trust for this edition.
Why DAF gifts run larger. The money is already committed to charity and the deduction is already taken, so it isn’t competing with a household budget. Chariot reports that digital DAF gifts to one nonprofit, March of Dimes, average about $1,100, against an industry-wide average of about $204 for credit card gifts. Chariot’s DAF Fundraising Report puts DAF gifts at roughly twenty times the size of non-DAF gifts, on average.
Each square is 1% of the asset base the 2025 Annual DAF Report uses for its payout rate. Measured against year-end assets instead ($64.60 billion granted, $327.87 billion held), grants come to about 20%. The DAF gift figure is March of Dimes’ gifts through DAFpay; the card figure is an industry average. Both come via Chariot, and neither is a sector average for DAF gifts.
Equity can be a large share of pay at senior levels in tech, which makes shares a realistic gift for donors who would otherwise give cash. A donor who has held shares for more than a year and gives them directly avoids the capital gains tax a sale would trigger, and can generally deduct the full fair market value, up to 30% of AGI.
The distinction that matters. RSUs are taxed as ordinary income when they vest, and the resulting shares generally start with a basis equal to their value at vesting. Freshly vested shares therefore have little built-in gain, and donating them doesn’t reverse the income already recognized. A gift of shares held a year or less can still avoid realizing the gain, but the deduction is generally limited to basis rather than market value. So ask which lot the donor is considering and when it vested or was bought.
The failure mode. Organizations lose these gifts by not being ready to receive them. If your brokerage transfer instructions aren’t published somewhere a donor can find them in under a minute in late December, the gift becomes cash or doesn’t happen.
Both lots are worth the same. Only the early one carries a large untaxed gain, which is what giving shares directly avoids.
Large tech employers match employee gifts, commonly one to one. Figures widely reported at the time of writing include Microsoft up to $15,000 a year, Disney up to $25,000, and Google, Apple and Salesforce up to $10,000. Treat each as an example to confirm with the employer, not a fact to quote to a donor: caps, ratios and eligibility change, and our sources are third-party databases rather than the companies themselves. Double the Donation’s current data, citing a Forbes survey of 250 large companies, has over 92% of surveyed companies offering some form of gift matching, with participation ranging from roughly 12% to 40% depending on how generous the cap is. It estimates $4 billion to $7 billion goes unclaimed each year.
Volunteer grants, in proportion. Microsoft and Apple pay about $25 an hour for employee volunteer time, and Google about $10. Salesforce gives seven paid days (56 hours) of volunteer time off a year, and Double the Donation lists its volunteer grants at about $10 an hour. That’s real money, but a full day of volunteering yields $80 to $200. Treat it as a way into the relationship, not a funding line.
Caps from Double the Donation’s employer entries, checked September 23, 2026. The 92% comes from a Forbes survey of 250 multi-billion-dollar companies. Confirm every cap with the employer’s own portal before quoting it.
Each square below is $100 of one year-end gift. Watch what happens to the squares as the gift moves from a pledge to money in your account.
Say a product manager at a large tech company plans to give you $10,000 this December. On a pledge card it’s one number. In practice it’s put together from several sources, each with its own paperwork, deadline and way of failing.
Each square is $100 of the gift.
Suppose this donor bought these shares years ago for $2,000. They’re worth $10,000 now, so $8,000 of that value is gain that has never been taxed.
If the donor sells the shares and gives you the cash, the sale realizes the gain. At a 15% federal rate plus a 9.3% California bracket, about $1,944 goes to capital gains tax, which leaves roughly $8,056 to give.
Illustrative rates. A donor’s real brackets depend on their income and where they live.
If the donor transfers shares they’ve held for more than a year straight to you, there’s no sale on their side, so there’s no capital gains tax, and the full $10,000 arrives. Your part is making the transfer easy.
Google and Apple are widely reported to match up to $10,000 a year, one to one. But no match exists until someone files the request, and the cap is shared across every charity the donor supports. For now it’s dashed: possible, not promised.
If a board member has pledged a challenge pool, this gift can release part of it. That’s real money, but it isn’t new money from this donor. “Your gift unlocks matching money a board member has pledged” is truer than “your gift triples,” and this audience tends to find it more persuasive.
Nobody files the match before the employer’s deadline. The gift comes through a DAF, and this employer doesn’t match DAF grants. The shares arrive on December 29, after a broker’s cutoff.
None of these is about persuasion. All of them are fixable in October.
Count the donor’s gift and money that’s already pledged. Treat the match as upside until the employer approves it.
1. A DAF grant isn’t a new deduction. The donor deducted the money when they funded the account, sometimes years earlier. Never tell a DAF donor their grant is deductible. It isn’t, and saying so damages your credibility.
2. Many employers won’t match a DAF grant. Policies genuinely split. Bank of America and CarMax exclude DAF gifts outright. American Express, Chevron, Dow and Edwards Lifesciences allow them. Look up the specific employer rather than assuming the layers stack.
3. Pledges and event tickets need care. DAF funds shouldn’t pay any part of an event ticket or membership that lets the donor or a related person attend or receive benefits, including the portion that would have been deductible had the donor paid personally. Pledges are more nuanced than the flat prohibition you’ll often read: IRS Notice 2017-73 sets out conditions under which a sponsor may make a grant that happens to satisfy a donor’s pledge, and sponsors differ on whether they will. In practice, avoid framing a DAF ask as a binding pledge at all, and check the sponsor’s own policy before December.
The One Big Beautiful Bill Act changed federal charitable deduction rules for tax years beginning in 2026. Three of those changes land directly on the donors this guide is about, and a lot of fundraising material still in circulation was written before them.
Itemizers can now deduct only the charitable contributions above 0.5% of adjusted gross income. A donor with $300,000 of AGI deducts only the portion of their total giving above $1,500.
Why it matters to you. The floor applies once to everything a donor gives in a year, not separately to each organization. Splitting $4,000 across eight charities deducts exactly the same as concentrating it in two, so this isn’t an argument for being one of fewer recipients. What it changes is timing: a donor who gives a similar amount every year clears the floor every year, while one who combines two or three years of giving into a single year clears it once and loses less. Expect more donors to bunch, and be ready to accept a larger gift in an on year.
Each square is $500. The dashed squares fall under the 0.5% floor, $1,500 a year at this income.
For taxpayers in the top 37% bracket, the value of itemized deductions is capped at 35 cents on the dollar. Where a gift produces a usable deduction after the floor and the AGI limits, that cuts each $1,000 deducted from $370 of value to $350, so $100,000 of deduction is worth $35,000 rather than $37,000. It isn’t the outcome for every top-bracket donor.
Why it matters to you. This is exactly the donor most tech-donor material targets. The tax argument for giving got slightly weaker for the highest earners in 2026. It didn’t disappear, and the capital gains argument for giving appreciated shares is untouched, but the framing needs updating.
A $2,000 difference. The gap grows with the gift, but it stays two cents per dollar deducted, about a twentieth of the tax benefit. Worth knowing, and not a reason to change your ask.
Donors who take the standard deduction can now deduct up to $1,000 (single) or $2,000 (married filing jointly) in cash gifts. About nine in ten filers take the standard deduction, so this may reach some of your donors, though high earners itemize far more often. A smaller version ran temporarily in 2020 and 2021; this one is larger and has no expiry date.
Why it matters to you. It excludes contributions to DAF sponsors. For your smaller and mid-level donors, this is an argument for giving to you directly rather than routing through a DAF, which cuts against the DAF-first advice you’ll read everywhere, including in section 06 of this guide.
State tax rules differ, so pay attention to where a donor lives before you talk about the tax side of a gift. The same gift of shares works out differently in Seattle, Austin or Chicago than in California. In Washington, the usual advice about DAFs also needs a caveat.
Giving appreciated stock produces two distinct advantages, and states treat them differently. Avoiding capital gains tax depends on whether the state taxes capital gains at all. Claiming a charitable deduction depends on whether the state allows one. A state can do one and not the other: Illinois taxes the gain but gives no charitable deduction, and Texas does neither.
So a donor in a state with no individual income tax still benefits from giving appreciated shares rather than selling them first. The state layer is simply zero, and the size of the federal layer depends on that donor’s own long-term capital gains bracket and whether the net investment income tax applies to them.
Approximate top marginal rates as of September 2026. Top rates apply only at high incomes and won’t apply to every donor. Verify current rates before using any of this with a donor.
Official state tax agencies and statutes where we could find them. Which states allow a charitable deduction: state treatment of charitable contributions.
Two different mechanisms, often confused. The first is that giving shares away isn’t a sale, so it triggers no Washington capital gains tax at all. That’s true regardless of where the recipient sits. The second is the state’s charitable deduction, which a donor claims against gains they realized elsewhere, and which applies only to gifts to organizations “principally directed and managed within the state of Washington.” National DAF sponsors such as Fidelity Charitable, DAFgiving360 (formerly Schwab Charitable) and Vanguard Charitable generally don’t meet that second test. A national DAF failing the deduction test doesn’t undo the first point.
That deduction is also narrower than most summaries suggest. For 2025 it applied only to donations above $278,000 and was capped at $111,000. Both figures are indexed, so confirm the current year’s numbers rather than reusing these. And a donor whose annual gains fall below the standard deduction owes no Washington capital gains tax at all. If you’re courting Microsoft or Amazon employees, raise the question rather than assuming either way, and leave the answer to the donor’s advisor.
A number of states offer no broad, generally available charitable deduction, including several that do tax income: Connecticut, Illinois, Indiana, Michigan, New Jersey, Ohio, Pennsylvania, Rhode Island and West Virginia. That isn’t the same as no benefit at all. Targeted credits for approved organizations or programs often still exist, and they can be worth more than a deduction: Indiana has no general charitable deduction, but its Neighborhood Assistance Program offers a credit of up to 50% of an eligible contribution. If you operate in one of these states, check whether your organization or program qualifies for a credit before telling donors there’s nothing.
Leverage is how far one donor’s gift can reach once other money stacks on it: an employer match, a challenge pool it releases, and shares given instead of sold. This estimator shows how large a gift could become, how much of it you can count today, and where it could slip. First, three donors whose gifts stack very differently. Then describe one of yours.
Illustrative These three donors are composites we made up to show the structure. They aren’t research subjects, and their figures aren’t findings.
Bars share one scale. Solid is money the donor is giving; dashed is a match that exists only once it’s filed and approved.
Five questions about one gift. The drawing updates as you answer, and the findings below are built from what you enter. It opens with an example donor already filled in, so change anything. Nothing you enter leaves your browser or is stored anywhere.
Each line is something that’s missing on your side or unconfirmed on theirs, with what to do about it.
Written for the 2026 season. Most tech donors don’t share one vesting calendar, so the most useful question is still the direct one: when does your equity vest, and does your employer’s match reset on the calendar year?
Brackets mark details only you can fill in. It avoids stating what anything is worth to the donor, which is their advisor’s job.
Most guidance in this space is written for organizations with a development department. If that isn’t you, doing a few things properly beats attempting the full program badly. Pick the one that describes your actual capacity, not your ambition.
You don’t need to change your mission or your work. You need to say what a unit costs, what it changes, and how you’d know if you were wrong. Pick a sector to see the same appeal written both ways, with each sentence marked by the job it does.
Illustrative The organizations, programs and figures below are invented to show structure. Don’t reuse the numbers. Use the structure, and put your own numbers in it. The template underneath is the part meant to be copied.
The basic route. Report what you actually observed, and be precise about what it is: how many people you served, what it cost, how many reached the outcome, over what period. Then say that you haven’t established what would have happened otherwise. This is completely respectable, and it’s where most organizations should start.
The advanced route. Add a comparison group and report a cost per additional outcome with its limits stated. It’s stronger, but it needs a credible comparison, matched time windows, a consistent rule for missing cases, and someone’s time to do it. If you don’t have that, the basic route told honestly beats an attribution estimate you invented.
Comparing your program’s cost to the cost of the crisis it addresses ($6,000 spent against $300,000 of avoided lifetime cost) is the most common construction in nonprofit fundraising, and the one this audience distrusts most. It assumes your program prevents the entire outcome for everyone it touches. Naming the share of people who would have got there anyway makes your number smaller and your argument far stronger. A funder who models things for a living reads an unhedged multiplier as either naive or dishonest, and neither gets funded.
Tech donors read a lot of short documents that have to survive scrutiny. A page or two with a defensible number will get further than a twenty-page annual report. We call it a one-page memo because that’s the discipline: if it runs past two, you haven’t finished editing.
Fictional example Riverbend Care Alliance and the Licensure Bridge don’t exist, and every figure is invented. We chose a workforce and health program on purpose: there’s no technology in it anywhere, and it still works on this audience, because the reasoning is what travels.
There are about 4,200 home care aides working in our region at a median wage of $16.40 an hour. Annual turnover in these roles runs near 48%, which means facilities are permanently short-staffed and families experience constant caregiver churn.
The obvious fix is licensure, which roughly raises the wage to $27 an hour and cuts turnover sharply. The reason it doesn’t happen on its own isn’t tuition, which is modest and often covered. It’s that licensure requires 480 clinical hours that are unpaid, scheduled during business hours, and impossible to complete while working the shifts that pay your rent. The binding constraint is income during training, not the cost of training.
We run cohorts of sixteen. Each participant receives tuition, a stipend that replaces lost wages across the 480 clinical hours, and a coach who handles scheduling conflicts with employers, which is the thing that actually causes people to drop out in month four.
Here’s why those three numbers differ. Across our 2023 to 2025 cohorts, 71 of every 100 people we enrolled completed licensure. Over the same period, 22 of every 100 comparable aides in our region reached licensure on their own, without us. So the honest attribution is 49 additional licensures per 100 enrollments, not 71.
That makes our estimated cost per additional licensure $19,200 rather than $9,400.
What this estimate doesn’t prove. Our comparison group is aides who applied but weren’t placed in a cohort, so they self-selected too. Treat $19,200 as our best estimate under that comparison, not a measured fact.
Three cohorts of sixteen, so 48 enrollments over 18 months. If those past rates hold, that produces roughly 34 completions and about 24 licensures that wouldn’t otherwise have happened.
Two things would send us back to you. If, in the funded cohorts, comparison-group licensure rises above 40%, our estimated cost per additional outcome passes $30,000, which is high enough that we’d want to compare it against alternatives before asking you for more. And if completion in the funded cohorts falls below 55%, the coaching model isn’t working and needs redesign before it scales.
We’ll report both numbers at month sixteen without being asked. If either threshold is crossed, we’ll tell you and recommend where to move the money.
Ask for the donor’s expertise before you ask for money. Invite them to test your numbers, send something useful with nothing attached, and only then make the case and the request.
Illustrative These use the fictional Riverbend example so the structure is visible. Replace the specifics with yours. Keep the sequencing and the restraint, especially the second message, which asks for nothing.
The second message asks for nothing. It’s the one people cut, and it’s the one that makes the sequence work. A message that delivers something and requests nothing is rare enough that it registers.
The unflattering number leads. Touch one volunteers the weakest point in your own analysis. That’s disarming to an audience whose day job is finding the weakest point in an analysis.
Twenty-one days, not fourteen. Four asks in two weeks reads as a sequence. Spaced out, it reads as a person. If they reply at any point, stop following the script.
Treat it as something to test, not a formula. Run it with warm prospects first, where a misfire costs least, and watch what actually gets replies. The third message has two openings depending on whether they answered. Pick the right one, because thanking someone for a reply they never sent is the fastest way to look automated.
There’s a widely repeated claim that tech giving peaks in mid-November because that’s when equity vests. It’s partly true and partly an artifact of copying. Here’s what actually varies.
Microsoft’s row shows its annual stock awards; on-hire awards vest on the hiring anniversary instead. Google’s row shows larger grants. Meta moves to two bonus payouts a year from the 2026 cycle. Nvidia’s dates come from an offer letter filed with the SEC; the rest come from advisory firms and news reports listed under sources. Individual grants vary, so confirm any date with the donor before planning around it. Vesting schedules compared ↗
Schedules differ by employer and by grant type, and the differences are large enough to matter. Meta vests quarterly, on February, May, August and November 15. Microsoft’s annual stock awards vest quarterly in the same four months, while its on-hire awards vest once a year on the hiring anniversary. Nvidia vests quarterly in March, June, September and December. Amazon’s grants are weighted heavily toward the later years, and those awarded at annual reviews typically vest in May and November. Apple vests twice a year, in April and October. Google vests larger grants monthly, which flattens the effect for most engineers.
Google moved its annual bonus from January to March starting in 2024. Meta has paid its annual bonus in March, and from the 2026 performance cycle it moves to two reviews and two bonus payouts a year. Amazon runs its annual compensation review early in the year, normally between January and April. Microsoft, whose fiscal year ends in June, pays annual bonuses in September. So a “Q1 bonus season” push aimed at a Microsoft donor is six months off.
The question that replaces the calendar“When does your equity vest, and does your employer’s match reset on the calendar year?” is a normal question in a cultivation conversation and takes ten seconds to answer. One answer from one donor beats any general schedule, and asking shows you understand how they’re actually paid.
Record it. Put the vest month and the match cap in your CRM as fields, not notes. Next year the outreach writes itself.
Many corporate matching programs run on the calendar year, and an unused annual allowance typically doesn’t carry over. Confirm both the year and the deadline with the specific employer rather than assuming. A donor with a $10,000 cap who has used none of it has, in effect, $10,000 of their employer’s money that expires.
Check the deadline, not just the cap. Submission windows vary. Microsoft accepts requests for gifts made in the preceding twelve months. Others cut off in January or February for the prior year. Confirm before you build a campaign on a date.
Brokerage transfers of securities aren’t instant, and late December is the worst possible time to discover that. Firms often need several business days, and some impose their own year-end cutoffs well before the 31st.
Publish a deadline of your own. Ask donors to start a stock transfer by around December 15, as an internal target. It isn’t a guarantee: completion depends on both brokerages, and some set earlier cutoffs of their own. Say it in November, not on December 20, and confirm the real date with your broker.
This is our suggested sequence, built from the deadlines in this guide. Written in September 2026 for the season immediately ahead. If you’re reading this later in the year, start at the phase that matches the calendar rather than trying to catch up. Tick tasks as you finish them; nothing is saved.
In our experience, most stock gifts that fall through fail on logistics rather than persuasion. This is the whole sequence, from being able to receive shares at all through to the acknowledgment. Work down it once before December and you won’t be improvising in the last week.
Matches are lost between the gift and the payment, usually because nobody owned the submission. Track them per gift rather than per donor, because a match follows the gift and dies on its own deadline. Fill this in for the gifts you already have, then work the blank cells.
The four questions that decide whether a match exists at all. Ask these before you spend time on a submission. Does the employer match at all, and at what ratio? Does this type of gift qualify, since many programs exclude DAF grants and some exclude stock? How much of the donor’s annual cap is already used, given the cap is shared across every charity they support? And what’s the submission deadline, which is often months after the calendar year ends but not always?
| Donor and gift | Employer | Gift type eligible | Cap remaining | Deadline | Submitted | Approved | Paid | Owner | Remove |
|---|
The column people skip is the owner. Assigning a named owner makes it much more likely a match actually gets submitted, and recording the deadline is what stops it being submitted late. In our experience those two columns are worth more than the rest of the table.
daoflo helps mission-driven organizations with brand strategy, communications, and AI strategy and adoption. If you want a second pair of eyes on your year-end campaign, or help turning any of this into your own numbers, we’re happy to talk.
Everything below was last checked on September 7, 2026. Where a figure comes from a third-party database rather than an original publisher, we say so, because those two things aren’t the same and guides in this category routinely blur them.