for broad access to our portfolio, which co-invests in every new company we invest in. Rolling Fund LPs also receive preferential access to limited-allocation deals.
We have a lot to be thankful for, so we’d like to start this update with a bit of gratitude. Thank you to:
OUR PORTFOLIO FOUNDERS. Thank you to all of our incredible portfolio Founders and CEOs who allowed us to invest in their companies, and have worked like crazy to build their dreams into realities. Particularly in the current market, where capital is abundant, Founders have endless choices of investors to take money from. Thank you to our 111 portfolio companies for choosing us, and for allowing us to ride along as a small part of their journeys.
OUR LPS. Thank you to all of our 2417 syndicate LPs and 94
LPs who have entrusted us with your capital. There are endless places to invest your money: dozens of asset classes, thousands of VC firms, and hundreds of syndicates and funds just on AngelList to choose from. Thank you for choosing us.
with our LPs, Portfolio Founders, and friends - for their assistance and expertise in identifying, assessing, and negotiating one or more of our investments. Thank you to our extended family of contributors who have helped us do more than we would have been able to on our own:
Aaron Samuels, Alex Kwon, Ali Jamal, Aline Lerner, Anish Acharya, Anthony Bertrand, Anton Borzov, Arjun Bhagat, Ashley Flucas, Bill Lynch, Brad Flora, Brian Nichols, Chad Byers, Chris Bergerud, Mo Amdani, Cyril Berdugo, Daniel Gould, David Adams, Derya Baris, Dileep Thazhmon, Dilraj Ghumman, Ed Roman, Evan Moore, Gilbert Gong, Gyan Kapur, Hiro Tien, Ivan Montoya, James Graham, Jared Fliesler, Jimmy Lee, Jon Hallet, Jonathan Wasserstrum, Kevin Moore, Kim Diocampo, Krishan Allen, Leonard Lynch, Marcus Stack, Matthew Dellavedova, Nathan Lustig, Neil Arora, Omar Haroun, Richard Lin, Roger Cawdette, Rohit Taneja, Ryan Li, Sajid Rahman, Sri Pangulur, Steven Coulis, The >Capital Team, Tommy Leep, Vaibhav Domkundwar, Will Babler, Zach Kruth, and Zak Holdsworth.
We would be 1/10th the size we are, if we didn’t have all of your support.
OUR TEAM.Since our last annual update, we’ve had 5 new team members join us: Thibault Reichelt, Chris Murphy, Declan Kelly, Alex Correia, and Sergii Zhuk - all of whom have been extraordinary and helped us expand our investing activity while keeping the quality high. Thank you to all of our team members for helping to build UV into something bigger, better, and more impactful than Peter would be able to on his own.
THE ANGELLIST TEAM, who makes the magic happen behind the scenes. Special thanks in particular to Rajeev Jotwani, our manager within AngelList, who is absolutely *AMAZING.* He really works wonders for us, and is instrumental to our operations at UV.
THE MARKET.We’d like to think we’ve made some good picks, but let’s be honest: the recent bull market has lifted everyone. Loved this meme: (
Our mark-to-market returns over the past 2.5 years have been better than expected, and extremely high for historical VC industry norms. But it’s not unique to us; most managers have seen exceptional recent returns. Time will tell if this continues, but we should at least be grateful for what we’ve had.
2. PORTFOLIO PERFORMANCE
For the first time, it’s not that early anymore. Our mark-to-market returns are excellent.
TVPI on dollars invested for 2019, 2020, and 2021 are 2.6x, 2.8x, and 1.1x. We have invested $29.8 M of principal, which has grown into $63.5 M of assets under management.
Before discussing our returns further, we have a few DISCLAIMERS:
This leans positive. This report is going out to a lot of people (>2400 LPs), and it’s not appropriate to share negative information about our companies so publicly (but the company-specific updates do share mixed news). Furthermore, VC is a game where only the extreme positive outcomes have a meaningful impact on portfolio performance - so let’s focus on those.
There may be errors here. All of the portfolio tracking is done by Peter and checked by the team with a spreadsheet. It’s very possible we got something wrong. If you notice an error, please let us know. We strive to make this as accurate as possible.
Startup valuation is tricky and complex. Different people do it in different ways. We explain our methodology in the appendix. You may disagree with it. You are welcome to run your own analysis and arrive at your own conclusions, using the raw data.
The only thing that actually matters is cash on cash returns. But in startup investing, those take a long time to materialize. The intent of this analysis is to estimate how we are doing in the short term, to evaluate if we are on the right track. You can’t take any of these numbers to the bank.
If you are a major LP of Unpopular Ventures, have invested at least $250k to date, and are willing to sign an NDA, we will share the complete portfolio data with you. Please submit a request via this form:
With that out of the way, we’ll break this into 2 sections:
Highlight a few investments that are doing particularly well.
Statistics about how our companies are performing.
Note that in past years, we talked about follow on investors too. You likely know by now that a LOT of our companies have been followed by Tier 1 VCs. We have at least 4 where we co-invested with or were followed by Sequoia, 3 with A16Z, and a ton more with Founders Fund, Initialized, First Round, Craft, Floodgate, General Catalyst, Tiger, Hedosophia, Index, CRV, Tribe, etc. - among many more that are widely considered Tier 1 as well. In the interest of brevity, we will stop there.
A. PORTFOLIO HIGHLIGHTS
Here, we’ll highlight a few of our companies that we think are making particularly good progress. Some remain confidential - so we have redacted identifying info.
[Confidential]: we have one company that we invested in in 2020, that recently signed a term sheet for a $50 M Series B at 50x the valuation we originally invested on, and a ~30x markup net of dilution. The individual VC Partner leading the round is a very famous Midas-list VC. Our investment here is now already worth >1x all the money we invested in 2020. [update in September: the
[Confidential]: we have another company that we invested in in 2020 that just raised additional capital at a 22x higher valuation, and a ~17x markup net of dilution. Thanks to our Partner Thibault Reichelt for leading this investment.
is a GitHub-like platform for business knowledge and documents. We were one of the biggest checks ($350k) into their first round, and they just closed a huge Series A on a 14x higher post money valuation from a major Tier 1 VC. Thanks to Kevin Moore for sourcing this investment, and to Brad Flora for co-leading the syndicate.
is an Open Banking API for Latin America. We originally invested in early 2020, were followed by Founders Fund on a higher valuation, and the company recently closed a $43 M Series B from a very famous Tier 1 VC on a 14x higher post money valuation. Thanks to our Partner Thibault Reichelt for sourcing and leading this investment.
is a rental car delivery service that makes it as easy to rent a car as it is to order an Uber. We invested in them 5 times on progressively higher valuations, and they are now closing a huge Series A on a 19x higher post money valuation than our original investment. Thanks to Gilbert Gong for sourcing this investment.
is a GPT-3-powered copywriting tool. Peter invested personally in their first round on $6 M post (before we had a rolling fund), we as UV invested further in their seed round, we were followed by Craft Ventures and Sequoia on a higher valuation [
is a “Super App'' for Francophone Africa’s 430 M population across 29 countries. We were one of the biggest checks into their seed round in mid 2019, were the first check into them during YC, and they recently signed a term sheet from a Tier 1 VC for a large Series A on a much higher valuation than our original investment. Thanks to our Partner Thibault Reichelt for co-leading this investment.
(formerly known as Cardea) makes infrastructure for modern hiring. We were one of the biggest investors in their pre seed, and they recently closed a seed round led by Backend capital on a much higher valuation. Thanks to our Partner Thibault for leading this investment.
(prev. KiranaKart) is building “Instacart for India.” We were able to get a small $25k check from our rolling fund into their seed round right after YC (couldn’t syndicate it because the round was too competitive) in Feb 2021. They already raised a Series A at a higher price, and are now in the midst of closing a huge Series B on a 17x higher post money valuation than our investment. Thanks to our Partner Thibault for leading this investment.
is building “cloud kitchens in LatAm.” We invested $400k in their seed round in 2020, and they just signed a term sheet with a top 3 VC for their Series A on a much higher valuation. Thanks to Ashley Flucas for inviting us to co-lead this syndicate with her.
We have A LOT more companies that are doing GREAT, but we are keeping this short in the interest of brevity. The above mentioned ones are just some of the more dramatic markups we’ve had so far. Also, a lot of our new team members’ investments are doing extraordinarily well, but not enough time has passed to see the markups yet (most companies take 1-2 years or longer to reach the next round). We expect to be able to share a lot more about our new team members’ successful investments in future updates.
We’ll lead with the numbers, with commentary further below:
Total to Date
Principal Invested (after fees)
Current Value (AUM)
# of Investments
# of New Co’s
# of follow ons in existing co’s
# that were fund-only or exclusive to fund LPs (fka “UV Preferred”)
TVPI (total value to paid in capital) - self reported (see appendix for methodology)
TVPI on Deals
TVPI on Dollars
TVPI on Fund
IRR (internal rate of return) - reported by AngelList (lags)
Net of Carry
There are no rows in this table
More color on this data:
We re-formatted this table from past reports for simplicity. If you’d like to see granular deal by deal portfolio movement - see the
We split TVPI by deals, dollars, and fund because we raise a different amount of money for each investment. “Dollars” means relative to cash invested. “Deals” means that if you invested exactly the same amount in every deal, this is your effective TVPI. “Fund” is how our rolling fund is tracking. We self calculate our TVPI, and you can find our methodology for calculating it in the appendix.
IRR: For the first time, enough time has passed for AngelList to begin reporting a meaningful IRR for us. AngelList currently puts us at a 76.9% gross IRR for the overall syndicate portfolio (63.8% net of carry).
IRR is tricky to calculate, so we won’t be computing it for each of our vintages as we do for TVPI.
Note also that AngelList’s data differs from ours, for a few reasons:
They don’t give us credit for deals that we co-led, but were hosted in another syndicate (but we do count them as part of our self-reported TVPI numbers).
They don’t mark to SAFE/note caps. We explain how we do this quite accurately in the appendix - and the intent is to have more-granular performance tracking. We have a lot of deals that have raised on higher note caps - but AngelList does not give us credit for them until they eventually do priced rounds.
They don’t update the numbers until the round is fully closed and wired (whereas we update when the term sheet is signed). As a result - AngelList lags us by a month or two for deals that are in closing. Similarly, AngelList takes a month or two to record new investments after they have closed.
We are off to a great start. But remember: we are only 2.5 years in on a >10 year journey, and it’s been a bull market. Some of our companies that are doing well will inevitably struggle, and some that didn’t take off right away will find their grooves and be extraordinarily successful. We are happy to see such high initial marks, but we still have a long way to go. We are hopeful that our strong trajectory will continue.
Framing these returns in the context of the broader market: 2.6-2.8x in 1-2 years is unusually good. Historically, 3x in 10 years was considered a top quartile VC return. As a hypothetical example, if we extrapolate the trajectory of 2.6x in 2 years out to 8 years (2.6^4) - you get a 46x fund. That would be extraordinary, but it also seems very unlikely. So being realistic, the rate of appreciation of our portfolio will most likely slow from here.
3. STRATEGY UPDATE
4 Sections Here:
Looking Back: what have we said in the past that came to fruition?
Team Building and Diversity
Looking Ahead: what should we focus on going forward?
Are there too many syndicates?
This section is written primarily by Peter (Founder + General Partner), hence the switch from “we” to “I.”
A. Looking Back:
There are a few specific things I talked about in past annual updates:
Our investment focus: “the best companies, off the beaten path.”
Incentives for deal sourcing and leveraging the power of a large community.
Syndicate as a firm: “the next Sequoia will be a syndicate.”
To the first point, what we invest in: “the best companies, off the beaten path” - we have said this in both of our past updates and it is clearly working for us. Our measurable returns are looking great so far. We are far from perfect, but we were able to invest in A LOT of companies early, at very attractive valuations, that went on to become much more valuable in a short amount of time (even have one 30x net of dilution already!). You’ll see examples of this in the Portfolio Highlights section above. We got in on those at great prices because we were able to build conviction before others.
Because this is working so well, we intend to keep doing it. Every company we invest in must be exceptional, with the potential to return >100x our investment - and we also want to be able to make the case that most are a little off the beaten path, or non-consensus in some way.
We were the first syndicate on AngelList to systematically share carry for value added services. This doesn’t seem so non-consensus now, because almost everyone is doing it. But I challenge you to find any mention of this idea anywhere, before our
It has been gratifying to see this idea be adopted widely, and even now productized by AngelList - who recently released new features for both syndicates and rolling funds that make it quick and easy to share carry.
I feel proud to have introduced this - and I think it has been a positive contribution to the AngelList community - making the entire platform stronger. The deal flow that we now see across the platform is the best it’s ever been, and I believe this is substantially due to enhanced collaboration across the entire AngelList community with the financial tools to reward it.
“Syndicate as a Firm”
This time last year, Unpopular Ventures was just me - and I theorized that syndicates might evolve to become more like firms, operating best as teams. I even went so far as to bet that “the next Sequoia will be a syndicate.” Since then, it has been interesting to see multiple syndicates experimenting in this direction - with Duro Ventures spinning out DVC and building a team, and Flight growing their team and operations as well.
We here at UV have been experimenting in this direction too. It started with Thibault and I joining forces - and it worked extraordinarily well. Thibault is amazing, has sourced many of our highest performing investments to date, and his network and deal flow have expanded the variety of companies we have been able to invest in - enhancing our portfolio, while keeping the investment quality high.
Based on how that went, we decided to take it another step further, inviting Chris, Declan, Alex, and Sergii to work with us. All of them are super smart, have impressive backgrounds, have sourced and led compelling deals, and we are thrilled that they agreed to join us. Big thank you to Thibault, by the way, for originally introducing all 4 of our additional Partners.
I’ll talk more about how we think about our team in the Team Building section below. For the purpose of looking back, this is still an experiment in progress - but I’m very happy with how it is going so far.
B. Team Building and Diversity
Our approach to team building has been simple: we welcome people who want to work with us, and who are willing to demonstrate that they have good judgment and can do good work.
Thibault, as one example and our first team member after Peter, was very easy to partner with because he sourced deals for UV, took the lead on numerous investments and wrote excellent memos with good due diligence. And then, many of those companies did well after our investment. As he continued to bring and “guest lead” deals with us - it was an obvious step to make the partnership official. Chris, Dec, Alex, and Sergii similarly sourced and/or guest led deals with us before we invited them on as Partners too.
The point is: our team building to date has been very organic.
But one thing I’m aware of, is that we haven’t done the best job of incorporating race and gender diversity on our team. I’d like to do better.
I would argue that we have done extremely well with other kinds of diversity on multiple levels - and likely better than 90%+ of VCs. Perhaps the most significant is that our Partners are on 3 continents and we invest globally - particularly in areas that are scary to most VCs, like Africa, India, LatAm, Pakistan, SE Asia, etc. Most VCs don’t invest outside their backyards, even if the color and gender of their team varies. Furthermore, our portfolio founders are themselves extraordinarily diverse; we’ve backed countless women and racial minorities, and well above average compared to most VCs.
But even so: race and gender diversity in the core investing team is top of mind for me, and something I’d like to improve upon.
So along those lines: if this is something you can help us with, and you are interested in working with us - please reach out.
C. Looking Ahead:
Our strategy is clearly working, so we will continue to do a lot of what we’ve talked about in the past. But there is one issue: a lot of the other syndicates have adopted our strategy/tactics, and a few have even gone so far as to plagiarize our work (notably our scout program guidelines, our old Preferred LP program, and some of our blog posts). As we shared above, most syndicates now leverage our idea of sharing carry for deal flow, which is certainly a positive for the overall AngelList community, but the wide adoption of this idea has made it no longer unique to us.
So the edge that our old strategy previously gave us is now likely gone, absorbed by our competition. Which means: it’s time to find the next innovations. And whereas we shared our ideas widely in the past, going forward - perhaps we will be better served by being a bit more secretive.
Having said that, here is one big thing that I think is worth sharing with everyone, and might benefit the overall AngelList community if more syndicates adopted it:
Metrics that Matter: Fundraising vs. Performance
One thing that is critically important for startups (including us - we are a startup too), is figuring out their “north star” metric. You tend to improve what you measure. If you can just figure out the one thing that matters most, measure it, talk about it incessantly with your team/customers/community, and execute against that one thing - that is frequently a path to success. It’s easy to get everyone aligned when everyone knows what the target is.
So what matters for building a successful syndicate and/or rolling fund? One thing I’ve noticed is that 80%+ of the communications from syndicate leads on AngelList are about how “oversubscribed” they are. Evidently, most leads believe the top thing that matters - the #1 thing worth talking about - is being oversubscribed. David Goldberg (GP at Alpaca VC) put it well:
I’ve been guilty of this myself in the past. And I know why everyone does it. When you send out a message on AngelList with the word “oversubscribed” in the title, you immediately get a flood of additional money into that deal. None of us like to think we are influenced by something like that, but the reality is that humans are herd animals and we are all influenced by social proof.
So it works. But it sucks. The “oversubscribed” fundraising marketing has turned into a
situation where everyone is doing it - but now our inboxes are polluted with 20+ emails per day about how oversubscribed, or almost oversubscribed, or half subscribed, or 5x oversubscribed everyone’s deals are. And the level of oversubscription is literally meaningless; many leads set their targets artificially low, so they can then come back later and announce how oversubscribed they are. It’s all about gaming our herd psychology.
You might have noticed that we don’t do it anymore. We stopped, for 2 reasons:
We don’t like how polluted the water has become, and have decided to lead with a better example.
We’ve decided to be more thoughtful about what really matters, and talk about that instead.
So what really matters here? Sure, it’s nice (for us) to raise a lot of money, because that means we earn more fees. But that’s not #1. In our opinion, the #1 most important thing is performance. Backing great companies at fair prices that have the potential to grow tremendously in value, and delivering outstanding returns to all of you, our LPs.
So instead of talking about fundraising - we have opted to instead focus on investing performance. Both in terms of how we communicate about new investments, as well as how we look back on them. Instead of hyping our deals with scarcity and social proof tactics, we will aim to do a better job of simply articulating why our investments are actually good. Social proof is often an element of that (startups tend to succeed when they can pull lots of good people around them) but that will never be the only part of our messaging.
And then afterwards, we track and look back on our past investment performance regularly, to evaluate how we did. Yes - this is hard to do well in VC, because the cash on cash returns take a long time, and the intermediate investment marking is messy and complex. To make it even harder, some companies see no price change for years and then get a 20x markup in year 5, while others sometimes get marked up quickly and then collapse a few years later. So it’s imperfect. But even so, we believe that this is the metric to focus on, measure it as well as we can, talk about it, and execute against it.
We measure our performance in detail, segment it by year and by partner, update it constantly, and talk about it internally all the time. You’ve likely noticed that it has become a major part of our quarterly updates to all of you. We believe that by focusing on performance as #1, making that the main thing we measure and talk about, over the long term - that will be the thing we excel at.
One example of how this helps to guide us is that we frequently see companies that we know will be popular on AngelList, but that we don’t actually think are great investments. A lot of syndicates pick those deals up, because they get the immediate gratification of raising a lot of money. We generally try to avoid deals like that because we are optimizing for investment returns vs. fundraising amount. We won’t always get it perfectly right, but this is one example of how we are trying to do better: by establishing clear values and sticking to them.
So that’s what we are doing: focusing less on fundraising prowess, and more on producing good returns for all of you, our LPs. I wouldn’t mind if other syndicates started behaving this way too. Our hope is that by focusing on returns as number one, we will continue to earn all of your trust, and actually make more money for everyone. And if we do well in that regard, perhaps the investment dollars will follow as a byproduct.
D. One Final Miscellaneous Thought: Too Many Syndicates?
One thing that has come up in many of my recent conversations with our LPs is about how many syndicates have appeared on AngelList in the last year, and if there are now “too many.”
I understand where this is coming from. Up until ~6-12 months ago, the number of syndicates and deals flowing through AngelList were few enough that it was reasonable for a serious LP to back every syndicate and review every deal. In fact, many of the serious LPs on AngelList strive to do that. I have done it myself.
But it now feels like we are at a breaking point. I can no longer keep up with everything, and I understand that most LPs can’t either. So the question is: is that a bad thing, and does AngelList now have too many syndicates?
I don’t think so; it’s just the next phase in AngelList’s growth and evolution. If you compare this to, say, Twitter, Facebook, or Linkedin - you wouldn’t try to follow every person on those platforms. That would be insane. Instead, you follow people you know, people you find interesting, and you unfollow people who don’t fit what you’re looking for on the platform.
So too should be the case with AngelList. Follow the syndicates you like, and unfollow the ones you don’t, until you’re happy with the quantity and quality of the content you receive.
In fact, it’s remarkable that many of us here were so early to AngelList that we could follow every syndicate. Imagine being that early on Twitter, where you could actually keep up with everyone there. But now AngelList is growing up, and it’s a good thing. There are tens of thousands of venture/angel deals that happen every year, and AngelList is getting access to more and more of them as time goes by. Maybe one day, there will be an AngelList syndicate in every single venture deal. But you as an individual aren’t going to ever be able to keep up with every one that happens. At some point you have to let go.
So the point is: if you’re feeling overwhelmed on AngelList, it probably just means it’s time to trim the number of syndicates you follow.
As an active LP on AngelList myself (>400 investments on the platform to date), I personally intend to cut the syndicates that spam me with excessive “oversubscribed” fundraising marketing. I’ll be sticking with the syndicates that make a clear effort to produce good investment performance, by 1) articulating why their investments are high quality (instead of only using hype), and 2) holding themselves accountable by reporting on the performance of their past investments.
4. PAST CONTENT
Here are some links to content we’ve written in the past.
Unfortunately, we’ve had 4 cases where other syndicate leads blatantly plagiarized our work. This isn’t against AngelList’s terms of service, and AngelList has not taken action against it. So we’ve taken enforcement into our own hands, and our policy is to immediately ban other leads we catch plagiarizing us.
We had one week in June in which 3 of our LatAm portcos announced mega VC rounds.
Methodology for calculating TVPI:
This is how we calculate the TVPI numbers you see above, and in the portfolio spreadsheet:
On all investments, we first incorporate the cost of the SPV or Rolling Fund admin fee - which ranges from 1% to 10%. As a result, “flat” investments are typically marked at 0.90x to 0.99x
We do not incorporate the cost of carry, because it varies between LPs.
For investments that have shut down, we mark them down to zero, or to the amount of partial cash that was returned.
For investments that are struggling (we currently have two that we classify this way), we mark them down an additional 50%.
For investments that have raised a subsequent priced round, we mark to the most recent share price.
For investments that have raised on a higher cap SAFE/note, we treat the most recent cap as the effective valuation, with some nuances:
Must be within reason. If a company raises on an outrageously high cap (or uncapped) with a huge discount - we don’t count it. It only counts if the discount is zero/minimal, and the cap used is reasonably close to what the valuation would be if it had been a priced round.
If it’s a post money cap, we treat it as the post money valuation. If pre, we treat it as the pre-money valuation.
We track and incorporate all of the dilution that would occur between the rounds, to the best of our ability. For example, if we invested on a $5 M post money SAFE, and the company then raises $2 M on a $12 M post money SAFE, we would mark that as a 2.0x - the delta between our post money (5) to the next pre money (10). If they then took $3 M more on an $18 M pre money cap SAFE, we would mark it up to a 3.0x → (10/5)*(18/12) = 3.
If a priced round is signed but not closed yet, we use similar math, but also add extra dilution to account for potential option pool expansion. If there has been no priced round for a long time, we add an extra 10% dilution. If there was a priced round very recently, with the option pool already refreshed, we incorporate less.
As soon as we receive a pro forma cap table, we revert to the official share price x number of shares we have.
Marking to Caps?
There is some debate about whether you can mark to a higher cap SAFE/note, or if you have to wait for a priced round before updating the valuation. AngelList, for example, only marks to priced rounds. However, other seed stage VCs we consulted say it’s common practice to mark to caps. In fact, the NVCA appears to believe it’s acceptable to use a cap as the effective valuation:
(source: NVCA SmartBrief on 5/27/20)
Historically, it made sense to only mark to priced rounds, because most rounds used to be priced rounds. And particularly for AngelList, with thousands of investments and a lean team - it’s most efficient for them to only mark to a share price. However, with the growing use of SAFEs, we are increasingly seeing early stage startups do 2-6 SAFE rounds over 2-4 years before eventually doing a priced round.
We even saw a $17 M “Series A” get done a few months ago by a tier 1 VC on a SAFE.
We certainly could wait for only priced rounds before updating our marks - but that means there will be no meaningful data to report for 2-4 years. We (and probably you too) would like to estimate our investment performance on a more granular level. We think the best way to do that is to mark to reasonable caps as rough markers of valuation, while calculating the effect as if they had been priced rounds.
If another investor is investing on a higher cap - they are legitimately investing on a higher valuation - and we think it’s reasonable to take credit for that.
Want to print your doc? This is not the way.
Try clicking the ⋯ next to your doc name or using a keyboard shortcut (