Is an ecard better than an email for employee recognition?

For most cases where the recognition is supposed to land as a moment rather than a record, yes. An email lives in the inbox graveyard and gets four seconds of attention on a normal Tuesday morning; an ecard arrives as its own artifact, often with the count of voices from the team, and the recipient actually keeps it. Email still wins when the artifact's job is search-archive (HR compliance, comp-committee record, formal documentation), but those cases are a small share of the recognition that gets sent as email by default.

When is email the right channel for recognition?

Three honest cases. When the recognition has to be in HR's email log for a future review or comp cycle. When the audience is not the recipient but a third party (a skip-level FYI, a cross-functional heads-up that someone carried a piece of work through). And when the recognition is operational gratitude with a brief written record (the Friday note that an on-call engineer averted an incident, the short acknowledgement of a customer-renewal mention). For anything that is meant to land as a moment the recipient will remember in six months, email is the wrong artifact.

Why do recognition emails so often fail to land?

Three reasons. The inbox is the channel where everything else also lives, so the recognition message competes for attention with deploy alerts, expense reports, and meeting invites and almost always loses. The recipient gives email about four seconds on a normal working morning and rarely re-opens it. And the email is a search result rather than an artifact, so it cannot be pinned to a desk, screenshot to a partner, or remembered in the small chemical-hit way that the recognition was supposed to produce. Effort is not the same as delivery, and email collapses the two.

Does an ecard count as real recognition, or is it just a digital greeting card?

If it is signed by the people who actually worked with the recipient, with specific lines that name what the recipient did, an ecard is real recognition in the sense the recipient will remember six months later. If it is a single generic cover with one boilerplate line, it is worse than no card at all. The format is not the recognition; the count of voices and the specificity of what each one wrote is what does the work. The ecard is the artifact that holds the voices. The voices are what makes it land.

How do I get the team to sign a recognition ecard so it does not feel empty?

Seed the card with your own first signed note before sharing the link, so the rest of the team has a register to match. Share the link in the channel where the team already lives, not in a fresh thread nobody is checking. Give a four-day window rather than asking for same-day signatures. And name the specific thing the recipient did in the seed note so contributors copy the specificity rather than the warmth. Cards that get fifteen specific lines beat cards that get thirty generic ones, and the difference is set by the first signature, not the prompt.

All articles

Recognition

Ecards vs Email for Recognition

Ben Holloway

May 26, 2026

Anders Holm hit five years at a small software company in Oslo on the morning of the 0.9 release. The recognition email from his manager landed at 9:02 a.m. local, three sentences long, sincere in the way those notes usually are, and by 9:07 it was buried under deploy alerts and a thread about a regression on staging. He read it for about four seconds. I know because he told me at a hotel conference kitchen six months later, when I asked him whether the company had marked the milestone in any way he could remember. He said yes, technically. He said he could not remember what was in the note, and was pretty sure he had never re-opened it. The card from the team, which they had also sent through a group ecard tool, was something he had on his desk, printed at home, for the better part of that year. The email was the record. The card was the recognition. The company had treated them like the same artifact, and they were not.

Start a group card in 30 seconds

Free. No signup needed.

Jump to section6 sections

Recognition has a job, and the inbox is not where it gets done

Most recognition that is supposed to mean something gets sent as email because email is the path of least friction for the manager doing the sending. You hit reply-all on the launch thread, you write a paragraph, you cc the skip-level, you ship the message in under two minutes. It feels like recognition because you spent your time and wrote real words. The problem is that the artifact you produced is mostly a record, not a moment. It lives in the same channel as the regression alert and the all-hands invite and the expense-report reminder, and the person you sent it to read it for four seconds in between two other things, on the morning of a release, on a Tuesday.

The job recognition actually has to do is to land, to stay landed, and to be the kind of thing the recipient can pull up six months later when they are second-guessing whether the work was worth it. An email does the first part poorly (four seconds of attention, then it slides), the second part not at all (the inbox graveyard is the inbox graveyard), and the third part only if the recipient is the rare type who searches their archive for kind things people said to them, which is mostly nobody. My position on this is that recognition which lives at the bottom of an inbox is documentation rather than recognition, and the two are different jobs that need different objects to do them properly.

What email actually does well, honestly

I want to be fair to email before I take the rest of it apart, because there are real cases where email is the right channel and an ecard would be the wrong one. The clearest is the formal compliance record: the promotion that needs to live in HR's email log because the comp committee will look it up next year, the kudos that ties to a performance review, the award notice the legal team wants archived. Email is correct for those because the artifact's job is to be findable in a search, not to land as a feeling, and ecards are pretty bad at the search-archive job (they live behind a link that may or may not be there in three years, which I will come back to later as a failure mode I have watched play out).

Then there are the heads-ups to people who are not the recipient of the recognition: the note to the skip-level that one of their reports just shipped something hard, or the FYI to the cross-functional team that the marketing lead carried the rebrand through legal review. These are status updates dressed in recognition language. The channel that matters is the one the people copied on it already check, so an ecard would be category-confused. Same goes for the time-sensitive operational note (Friday afternoon, on-call engineer averted a real incident, you wanted it in writing before the weekend) where brevity itself is the point.

One sub-case worth pulling out, because it cuts the other way. There are also times when you are not collecting voices from a team and are sending recognition from yourself alone. Even then the ecard often beats the email, for reasons that are about the artifact rather than voice count. A one-to-one ecard from a manager to a report on a quiet Tuesday in February, with no team involvement, is a different object than the email with the same text. It has a cover, it has a delivery moment, it opens into a slightly different mental register than the inbox does. The recipient knows, while reading, that you took the extra two minutes to use a different channel, and that small signal is most of what recognition needs to do. The email from a manager reads as part of the manager's job; the card reads as a choice the manager made.

So if your recognition fits the compliance record, or the third-party heads-up, or the brief operational note, send the email and stop reading. The rest of this piece is for the ninety percent of recognition that does not fit any of them and is being shipped as email anyway, because the sender confused friction-free with channel-fit.

Where ecards actually do work email cannot

The argument for an ecard at the moments that actually matter is not that it is prettier or has a cover image. It is that for the cases where recognition has to land as a moment rather than a record, a few things happen at once that the inbox cannot reproduce, and I want to point at the two or three that have mattered most in the recognition I have actually watched land or fail to land.

Start with the count of voices, which is the obvious one but more important than people give it credit for. A team ecard for a work anniversary signed by twenty-three people does work no email can do, because the head count itself is part of the gift. The recipient sees, in a glance, the number of colleagues who took two minutes to write something specific. The email from one manager, however warm, cannot deliver that number; twenty-three separate emails is a clogged inbox where nobody reads past the third one. The count only works if the format collects the voices in one artifact.

Then there is the writing itself. People just write differently in a card than they do in an email. The card invites a sentence that is about the person; the email invites a paragraph about the project. I have read recognition emails that were excellent paragraphs about what someone built. I have read card signatures for the same person that were three lines about what it felt like to sit next to her on the day of the big launch and watch her stay calm. Both have a place. Only one is recognition in the sense the recipient will recognise as recognition six months later, when they are pulling something up to remind themselves the work was real.

And then there is the artifact problem. The card lives somewhere outside the inbox. People bookmark it. They take a screenshot and put it in their personal slide of nice things people have said about their work. Some print it and put it on the corner of the desk; some email it to their parents (genuinely, this happens more than you would think; my own mother has three of mine in a folder somewhere). The email, by contrast, gets buried by Friday, archived by quarter-end, and surfaced again only by accident in a quarterly inbox-zero session that the recipient runs while trying not to think about anything. The card is an object you can return to. The email is a search result you will not search for.

The two times I picked email and should not have

A short list of my own failures, because abstract advice on this topic is much less useful than a specific botched send.

The first was a recognition email I sent in late 2022 to a contractor named Sloane, who had carried our entire customer-success function for the back half of the year while we were searching for a head of CS. She was technically a six-month contract; she had stayed for fourteen months by the time of this email, because we kept extending and she kept saying yes. The email was four paragraphs, written carefully, sent on a Thursday morning at the end of her fourteenth month. She wrote back the next day, warmly, in two paragraphs. Three weeks later, in a 1:1, I asked her how the recognition note had landed. She paused, smiled in the way you smile when you do not want to be a problem, and said she remembered I had sent something nice. She did not remember what it said. I had sent a record and had thought I had sent a moment. The card from her seven CS colleagues, organised by one of them on a group ecard platform without my involvement, was the artifact she still had open in a tab. The email had done what emails do.

The second was a recognition email I sent in 2023 to a designer named Mehmet who had carried us through a brand refresh that took six months and three near-failures. I copied his manager, I copied my skip-level, I wrote a real paragraph naming three specific decisions he had made that I thought were excellent. The send felt good. Mehmet replied with a polite thank-you the next day. The brand launch happened two weeks later, the team had a small in-person thing, and at the dinner one of his colleagues mentioned that 'someone should have done a card'. I felt the heat in my face. I was the someone. The email was real recognition in the sense that I had meant it. It was not recognition in the sense that it produced any artifact Mehmet could hold up at the dinner, or screenshot to his partner, or pin to the corner of his monitor. The card his colleagues retroactively organised that week was the thing that lived for him. My email was a search result in his archive.

Both mistakes had the same shape, which I now recognise: I conflated taking the time to write real words with actually delivering recognition. The first is a virtuous input. The second is a different artifact entirely, and what survives in the recipient's memory at the end of the year is mostly about the artifact rather than about how hard I thought.

I should admit the rule has a real exception, because I keep finding it inconvenient. There is a developer I worked with on a hard release in 2021, and in his case the recognition that stuck was a long email I sent on a Sunday night in late November when the launch had been a near-failure and I was still wound up about how he had held the line through it. There was no card because the team was three people and one was on PTO. The email was four paragraphs, the longest I have ever written for that purpose, and he forwarded it to his father, who is a retired engineer; his father then printed it and stuck it on the noticeboard in his garage workshop. I learned this two years later when the developer mentioned it in passing. That email outlasted half a dozen team cards I have organised since. I do not know exactly what to do with this except to flag that the channel-fit framing is mostly right but not absolutely right, and that occasionally an email at the right hour with the right person on the other end will end up on a noticeboard in a garage in suburban Massachusetts.

The hybrid, and a quick five-second test before you ship anything

The honest answer to the binary, for a lot of the harder cases, is to do both. Send the email for the record, the compliance trail, and the skip-level audience. Send the ecard for the moment, the count of voices, and the artifact the recipient will keep. The hybrid is the right call for promotions where the comp committee needs the email and the recipient also deserves a moment that is not a search result. Work anniversaries past the first one. Any time you are about to send a long recognition email and notice yourself thinking that the email is doing two different jobs (and if you find this is happening to you a lot, the line-level mechanics of writing the card half of the hybrid are basically what thank-you messages for a coworker covers, with the same specific-line-over-generic-warmth idea I keep coming back to). Split the jobs. Send the email for the record. Send the card for the recognition.

This is not double work, exactly. The email is shorter when you know the recognition is also coming through a card. The card is more focused when you know the operational record lives elsewhere. I have sent the hybrid somewhere north of a dozen times since I figured this out and have not once regretted it. The cost is about four extra minutes of the organiser's time, and the lift in what the recipient remembers six months later is, as far as I can tell, the difference between 'I think she said something kind' and 'I have it on my desk'.

If you still cannot decide between channels for a specific recognition you are about to send, the test that works for me is small and quick. Imagine the recipient opens your message on the morning of a normal working day with eighteen other things in their inbox, gives it the five seconds of attention people give email on a normal Tuesday, and then never re-opens it. Is the recognition accomplished? If yes, send the email. If no, the email is the wrong artifact for what you are actually trying to do, and you should either send a card or, more commonly, send both. This test cuts through most of the bad recognition I have sent, because in most cases the honest answer is no; the thing I wanted to give the person was a moment, and I gave them a record instead because the record was faster to produce. I learned this slowly.

Turn it into a group ecard the team actually signs

For the recognition that needs a count of voices, what I would actually do is just send a group ecard, built in whatever channel the team already lives in, signed by the colleagues who actually worked with the recipient, scheduled to deliver at the moment that matters. A group ecard with multiple signers collects twenty or thirty short specific notes in one artifact in the four-day window before the work anniversary, the launch debrief, or the quiet promotion announcement. You can create a card online in a couple of minutes, share the link to the team channel with one sentence of guidance, and let the signatures fill in on their own time.

One thing that matters more than people think: seed the card with your own first signed note before you share the link. A specific opening signature sets the register for everyone else and you will get fewer 'thanks for everything' submissions back. The longer walkthrough at how to create a group ecard covers the seed-it-first move, the PIN-protected delivery, and the four-day window. I should also flag the failure modes I have seen, since I promised earlier I would: an unsigned cover that lands as a single name with a blank space behind it because nobody clicked through; a cover image that doesn't render in some corporate email clients (Outlook on locked-down enterprise machines, occasionally); and the genuine archive person who actually does prefer email because they keep everything in a Gmail folder and an ecard URL is not something they can file the same way. I have run into all three, the first more than I would like to admit.

For the broader program question, by which I mean why most recognition programs fail and what small rituals actually move team behaviour, employee recognition ideas that actually work covers the program-level version of what this article only covers at the artifact level.

One last thing, off-topic and probably just for me. The kitchen at the conference where Anders told me about the recognition email he could not remember was on the second floor of a hotel I have been trying to picture again as I write this and am mostly failing. I could not tell you the colour of the carpet now. I remember a fluorescent light that buzzed audibly in the corner above the milk fridge, and a row of motivational posters in three languages on the wall by the door, one of which was in Norwegian and which I could not read at the time and still cannot. I remember the smell of the coffee machine more than the look of it. I think it took about a minute and a half per cup but it might have been longer; I was hungry and I had not eaten and time was running long. We talked for around eleven minutes, give or take, and the only thing he said about the email I have not been able to forget in the four years since is that the print of the team card had a small smudge on the lower-left corner from where he had handled it too many times. He said it the way you say a thing you have noticed about an object you are fond of, not as a complaint.

The email was archived in a folder he had not opened. The card had a smudge on the corner from a year of being on his desk. That is most of what I wanted to describe in the two-thousand-something words above this paragraph. The smudge sits there on the corner. The Slack message about the regression on staging scrolled off the screen the same day.

And then we walked out of the kitchen and back into the conference. Anders went to a talk about distributed tracing, or maybe it was something about CRDT merging in collaborative editors; I genuinely do not remember. I went to the wrong room twice and ended up in a session on compliance for fintech in the Nordics, which is not my field at all, and stayed for forty minutes anyway because the speaker was funny and I did not want to admit to the people around me that I had wandered in by mistake. The parking lot still had the three buses in it when I left. One of them had its hazards on. I do not know where any of them were going. I have not thought about that kitchen more than three or four times in the years since; I thought about it again this week because of this article, and because the print of his team card came up in a conversation with a friend about whether the photo printer on her desk was worth the counter space. She said yes. I do not own a photo printer. I think about it sometimes.

Frequently asked questions

Is an ecard better than an email for employee recognition?: For most cases where the recognition is supposed to land as a moment rather than a record, yes. An email lives in the inbox graveyard and gets four seconds of attention on a normal Tuesday morning; an ecard arrives as its own artifact, often with the count of voices from the team, and the recipient actually keeps it. Email still wins when the artifact's job is search-archive (HR compliance, comp-committee record, formal documentation), but those cases are a small share of the recognition that gets sent as email by default.
When is email the right channel for recognition?: Three honest cases. When the recognition has to be in HR's email log for a future review or comp cycle. When the audience is not the recipient but a third party (a skip-level FYI, a cross-functional heads-up that someone carried a piece of work through). And when the recognition is operational gratitude with a brief written record (the Friday note that an on-call engineer averted an incident, the short acknowledgement of a customer-renewal mention). For anything that is meant to land as a moment the recipient will remember in six months, email is the wrong artifact.
Can I send both an ecard and an email for the same recognition?: Yes, and for promotions, work anniversaries past the first, and any recognition where HR needs an archived record, the hybrid is usually the right call. The email is shorter when you know the recognition is also coming through a card; the card is more focused when you know the operational record lives elsewhere. Each artifact does one job well rather than both badly. The extra cost is about four minutes of the sender's time and the lift in what the recipient remembers six months later is substantial.
Why do recognition emails so often fail to land?: Three reasons. The inbox is the channel where everything else also lives, so the recognition message competes for attention with deploy alerts, expense reports, and meeting invites and almost always loses. The recipient gives email about four seconds on a normal working morning and rarely re-opens it. And the email is a search result rather than an artifact, so it cannot be pinned to a desk, screenshot to a partner, or remembered in the small chemical-hit way that the recognition was supposed to produce. Effort is not the same as delivery, and email collapses the two.
Does an ecard count as real recognition, or is it just a digital greeting card?: If it is signed by the people who actually worked with the recipient, with specific lines that name what the recipient did, an ecard is real recognition in the sense the recipient will remember six months later. If it is a single generic cover with one boilerplate line, it is worse than no card at all. The format is not the recognition; the count of voices and the specificity of what each one wrote is what does the work. The ecard is the artifact that holds the voices. The voices are what makes it land.
How do I get the team to sign a recognition ecard so it does not feel empty?: Seed the card with your own first signed note before sharing the link, so the rest of the team has a register to match. Share the link in the channel where the team already lives, not in a fresh thread nobody is checking. Give a four-day window rather than asking for same-day signatures. And name the specific thing the recipient did in the seed note so contributors copy the specificity rather than the warmth. Cards that get fifteen specific lines beat cards that get thirty generic ones, and the difference is set by the first signature, not the prompt.