Slashdot Mirror
Ask Slashdot: How Do You Sell an Algorithm To Venture Capitalists?
dryriver writes "Dear Slashdotters, We are a two man crew who have spent almost three years developing a video processing algorithm that 'upgrades' the visual quality of digital video footage. We take video footage that is "of average quality" — think an amateur shooting on a cheap digital camcorder or on a smartphone camera — and use various mathematical tricks we have developed to make the footage look better — optically sharper, better lit, more vivid colours, improved contrast, enhanced sense of three-dimensionality and of 'being-there realism.' In about a month, we will be presenting our algorithm to some venture capitalists. We have the obligatory before-and-after video demos prepared for this, of course. But there will also be a short PowerPoint presentation where we explain our tech in some detail. Now here is our main question: What, in your opinion, should we — or indeed should we NOT — put in the PowerPoint presentation to impress a Venture Capitalist? Should we talk about how we developed the algorithm at all — what kind of R&D and testing was involved? Should we try to walk the VCs through how our algorithm works under the hood — simplified a bit for a 'non-engineer' audience of course? Or should we stick to talking about market potential, marketing strategy & money-related stuff only? If you were in our shoes — presenting a digital video-quality improvement technology to professional VCs — what would and would you not put in your PowerPoint? Any advice on this from Slashdotters with some experience would be most welcome!"
how it will make them money.
Know your audience, that's for sure. As far as how much detail to let them in on, send me the copy of the details and I'll get back to you.
I would focus your presentation on how the VCs would get ROI from your technology, in other words: what's in it for them. For example, what do you see being the primary use for this technology? Who would it benefit and how? How is this an improvement over what is already available? Etc.
That's the way it's going to end anyway. At least this way you save time, money, hassle, and the anguish of actually thinking they give a shit about your success.
VC: "Ideas are like assholes, everybody has one and I'm going to get as much money out of selling yours as I can."
Don't give your future away by taking money from VCs for this. It sounds like you have enough already to write a small standalone processor for video files and charge money for it. You give up a lot - everything if you're not careful - when you do a deal with VCs. If you can build a company with any profits at all on your own, do it.
Do not go into too much internal detail. Focus on visualization of outputs. Make a powerpoint with these visualizations that walk through some real world problems. Get your elevator pitch down pat. Profit!
1. Do you have Patents on the Technology?
2. What kind of money are people/corporations willing to pay for this type of video enhancement?
3. How big is the market? (e.g. Can you deploy the technology in a smartphone app, or does it require specialized hardware)
They are venture capitalists, their goal is to make money and therefore you need to show a how and why what you are doing will make them money. So you need to show :
a) What your doing will sell and sell well, hopefully you have the market research behind this.
b) Your ideas are original, can't be easily copied by competitors or you have the patents/copyright behind it and you should be able to articulate this.
c) a good demo and complete openness when it comes to them asking questions, you need to be able to convince them you are worth the bet and nothing turns an investor off faster than you being less than forthcoming.
The deep technical details comes a distant second to the business aspects.
You don't sell technology to venture capitalists, you sell a business plan with well done research and projections.
Or, alternatively, you should start by licensing the technology to those interested first (figure out who) and create a steady income from there.
Or create a product that end users might find useful directly, like an iOS app.
One example of successful business model based on an algorithm is elastique, which is used in pretty much all major audio and DJ apps.
Hope this info is useful!
It'll sell really fast if you include those features.
As people has said, know your audience and what they expect.
Watch a few episodes of “Shark Tank”. Its entertainment so don’t take it as gospel. But it does show a lot of people doing rookie mistakes. Such as talking about how big the market is and if we can only get a sliver.
Also, focus on making a clean presentation.
Make sure all of the programs you will be using are up and running prior to the presentation – don’t start the programs in the middle – you are just asking for problems. Turn off everything else – in particular things that pop up.
Little errors & inconsistences make it seem you have a half-baked product.
There is a fine line between confidence and arrogance, between insecurity and honesty (I am assuming your product is not 100% perfect yet). . Find that line before you go in.
Tell them broad strokes, not enough to figure it out. Remember that they're called Angel Investors because they can swoop in, take your idea, and do it themselves with their own money. Otherwise, most advice here is solid. They want to see how you develop it, what the exits are, and how much they can reasonably make.
If you can explain it in ways they can understand then its probably not worth much and if you do explain it they can share it with others.
You're really lucky you have an algo where they can see the effect, leave it like that.
They want a cash flow over time, which means:
1: Patent protection
2: A version 2,3,4,5, read up on Dolby Studios, they started decades ago doing the same thing but for hissy audio tape, still going N versions later on totally different media.
3: Get the word "mobile" in this. VCs are obsessed with mobile currently, I assume this will make phone pictures better ?
4: They want an exit strategy, is this going to be sold to join people's patent armouries or a firm they can float ?
5. Have you talked to Google, this sound like just what they might want for YouTube and to stop others getting it.
Dominic Connor,Quant Headhunter
Your discussion should mention that it is patent pending. Otherwise they will probably not be interested.
Here's my take. As others have stated VC's are in this for money, thus you show them the business case and that incudes explaining the barrier to entry - why another 2 bright guys, suitably funded couldn't easily reproduce your efforts. Now my real advice is this: VC's are not interested in small play's, they want to put a significant chunk of money in, and take a large multiple on that back. What that means is they are interested in large $$ opportunities that may take some time to come to fruition ...and that involves plenty things that can go wrong along the way leaving you with nothing to show for it. You are 2 guys, you already have something you think is very impressive to show, you don't really need lots of capital to continue development, it's really just your time. Go shop this to the incumbents in the applicable spaces looking for an immediate sale, Adobe/Apple in authoring S/W, google/vimeo on the backend, go-pro/contour/canon/nikon etc etc on the consumer H/W side. What you are looking for is enough money for the 2 of you to have a nice 7 figure bank balance..which means sell it, go work for the buyer for a while, take the money now whilst there is no obvious competition or the market changes and minimize the risk. Don't hold out to be a bizzilionaire, the chances not matter how good it is are vanishingly small.
I've seen a whole lot of video enhancement 'algorithms', boxes, programs, etc. Not one of the press-button-get-video variants have actually improved video quality, and almost universally they just make things worse; more often than not, they just shove up contrast and saturation and add an unsharp mask, but some are genuinely innovative in their uglyness (e.g. the dreaded WarpSharp, Q-Tec's BD butchery, etc). The vast majority of 'easy to use' variants, with a few sliders to move about or checkboxes to flip, are equally ineffective.
Do you have any examples of your 'algorithm' that show it to be something other than run-of-the-mill?
Tell the VCs how you are going to make them rich.
Just a thought if a lot of people are going to benefit from your algorithm then you could think about having a kickstarter campain to open source the code. With the amount you are looking to get from the VC's. Maybe with stretch goals of a iPad and Android version. You'll get money up front and might also be able to license it to somebody that doesn't like either (lgpl or gpl) depending on the strategy you choose. You could try having the opportunity of corporate sponsorship and trying to contact Google they are probably going to love having something that makes Youtube videos look better. It would trow some money your way because we need more funding of R&D through crowdsourcing. And remember to have some accessories for people like T-shirts.
There would be exactly one powerpoint slide that goes something like this:
1. Take the consumer's crappy video.
2. Run it through our patent pending thingy and make it look nice.
3. Profit!
(1) The team is most important. Have you assembled a team with the necessary experience to convert an idea into a successful product. VCs mostly invest in people, not so much ideas. Ideas are plentiful, people that can ***deliver*** are not.
(2) The idea is of secondary importance, but it must be something that can be protected by ***patents*** and/or other intellectual property. It does not matter how good the idea is if it is something that can be replicated by others.
We have the obligatory before-and-after video demos prepared for this, of course. But there will also be a short PowerPoint presentation where we explain our tech in some detail. Now here is our main question: What, in your opinion, should we — or indeed should we NOT — put in the PowerPoint presentation to impress a Venture Capitalist?
Give them an overview of the process. High light the points that can be protected by ***patents*** and other forms of relevant intellectual property.
Should we talk about how we developed the algorithm at all — what kind of R&D and testing was involved? Should we try to walk the VCs through how our algorithm works under the hood — simplified a bit for a 'non-engineer' audience of course?
They will not care beyond determining that the algorithm is in fact ***your creation*** and not related to previous employment or other potential conflicts.
Or should we stick to talking about market potential, marketing strategy & money-related stuff only?
They both care and don't care about this. They care in the context of determining if you are being realistic. Have three sets of numbers. One covering the scenario if things go as planned, another more pessimistic one, and another more optimistic one. The pessimistic one should not lose money.
That said, they will not trust your numbers. They will do their own research and do their own numbers. Do not let this deter you from doing the best job you can on your numbers. You are being ***tested***.
VCs have different contractual terms, but in the end they want to see at least a 4 to 1 ROI. As others have said, you need to tell them how it will make them money. Do you expect to have a software company and sell an app? Will you license this tech? If so, who is your target. Does this take time - meaning if Samsung was a licensee would this algorithm work in real time on a phone? Will someone have to transfer video to a PC and then post process? If it is the latter, you really need to talk about who will buy it. Have you done any market research? You have to convince the VCs that this will make them money - so you should have a solid idea of who your customer is, and if it will sell.
Feature, Advantage benefit.
For each feature the software has, how is it an advantage and how does it benefit the customer.
Also the one thing VC's really want is data to back up your claims of sales. What market research have you done, Who will buy this product? How many will buy the product? Is it a need, want or nice to have? What are your costs? Is there customer support involved? Is the a mass market or niche market? How will people find your product?
The bottom line for a VC is "How am I going to have a reasonable likelihood of making money on this product?". Do not go into technical detail on how things work. At best this will bore them; at worst it will make them feel stupid. Most VC's don't care how it works; they care that it works well and will make them money. Be prepared with technical details because you may be tested to see how well prepared you are.
I agree with the rest of the posts that it is all about the business case. One critical component is how easy it will be for someone else to duplicate if the business is successful. Patents can be worked around and expose key elements of the technology to examination, so it is OK for someone else to accomplish the same results with a different mechanism. Technically algorithms can't be patented, but you can patent it, if you say its part of a system. So it may be more valuable if you have something so advance that no one else is capable of reproducing it. If you have years of research with top level people then you should mention it in the presentation. This includes marketing research etc. Being first to market is also important, and you will be asked how you know someone is not doing the exact same thing right now and has it ready to launch. Anything that makes your work unique will help answer the question.
Don't be afraid to confuse them - just be sure that you prove the benefits of your idea first. A hands-on, real-time, first person demonstration of before and after would probably work best.
Once you've shown them that it works, launch into a short thesis of the technology, and don't be afraid to use technical terms and equations and leave them shaking their heads - this is a good thing because they will walk away saying things like "I'm really impressed."
This is because people in general, including vc's, devalue anything that is explained in simple, understandable terms, thinking "if even I can understand it, it must not be all that impressive."
Just make sure you prove that it works first.
Are you telling me you can't acquire 80's pr0n, run it through this algorithm, and then make a bundle reselling it. Recent reports say this is more profitable that Netflix and prime. This should give enough funding to do whatever else is wanted.
"She's a scientist and a lesbian. She's not going to let it slide." Orphan Black
I've worked directly with a lot of VCs and private equity guys.
First question you have to ask is "do I really want VC money?" Make no bones about it, VCs will scalp you. They will take control of the company and there is a good chance you will be pushed out. (No you will not retain control ala Mark Zuckerberg in all likelihood) Furthermore the cost of capital from a VC is VERY VERY high. They expect a big return on their investment. If you can fund your company with Angel investors, loans, friends and family, etc, then do that. You'll come out better in the long run most likely. Most people that think they want VC money really don't. I know it's a bit of a cultural thing in Silicon Valley but even the VCs will tell you that you don't want their money if you don't need it.
If you decide that VC money is still the best way to go then yes they will want to see the technical details of your product. They will do considerable due diligence regarding it and they will call in experts they trust to look it over. However you don't need to get carried away with technical details in the initial meetings. What they are going to be looking for is a business model around the technology that they can develop. VCs aren't typically going to want to develop the business themselves - they want to invest in people who are going to develop the business. They may bring in their own people to help or take over if needed. So what you need is to be able to explain very concisely who you are and why you are worth investing in, why this technology matters, what the market opportunity is, who the competition is and how you will overcome them, and how the VCs will be able to get a return on their investment. You should be able to say all of the above in 60 seconds or less.
Don't assume your technology is unique from a business standpoint - it probably isn't. There will be competing products out there and you'll need to explain why your technology is special and more importantly why you and your team are especially positioned to take advantage of whatever market opportunity you see.
Frankly VCs will be less interested in the technical details than in is your business plan and especially in the team you have around you to execute said business plan. VCs really don't invest as much in technology as they do in the people who are trying to bring that technology to market. They are investing in YOU more than anything else. Tell them about your leadership team. Show them your team has a track record of bringing technologies to market successfully. Guys like Elon Musk can raise money for almost anything because of their track record. If you don't have a track record of building successful businesses, get people on board who do BEFORE you try to talk to VCs. Get a team of advisors who can poke holes in your business plan and help you get on the road to success.
Finally, VCs tend to specialize. Odds are most won't be interested in what you have. Kind of like publishers you have to look around for the right one and you'll probably get a lot of doors (politely) slammed in your face. The VC community is a small one and they tend to all know each other. Don't be rude or talk smack about anyone else because word will get around.
As others have said VC's care about money, but the money your company can make is a direct function of the value of your algorithm (and also team etc.). Other commenters have talked about business plans, market reserach etc. but this is really secondary. What is of primary importance is what is your algorithm in itself, what does it do? Nothing you have claimed so far makes your algorithm stand out as something of value. If I were a VC I would ask you the following: Does your algorithm automatically fix the video with no input parameters needing to be speicifed? If it needs parameters, how is it different to composing a number of conventional video filters? Also do you have videos comparing the "best" current algorithms applied to video, with your own? Remember you are selling your algorithm, not the video editing software that could theoretically be built using your algroithm, as such software could also use other algorithms too, so that is not your competitive advantage.
Do not give them technical details, a before and after video will be enough. If you give them details they may just try and rip you off. If you give them details make sure they are not complete enough to reimplement your work.
Ideally don't deal with Venture Capitalists at all, these people are only interested in taking as much of your business away from you as they can get away with. They can't be trusted.
It sounds like what you really need is a sales guy to go out and sell this technology under license, not a venture capitalist.
They could do that. But they need to know who will buy it, how much they'll pay for it, and how many customers they'll have.
A VC doesn't care what it is. It could be a CD of killer tunes, or a skyscraper. What they care about is whether they'll actually make money from it. To convince them that they will, the pitcher needs to show them the market.
Sure you'll cover the market & the work, but the first thing a clueless VC wants to know is what the competition/alternatives look like. Not in-depth company analyses, but what an average consumer may find & try in 1/2 hour and how poor their results will be.
Also, BusinessWeek has a recent presentation that won major VC buy-in. It's where I got the above tip from & there's probably more to mine from it.
Science & open-source build trust from peer review. Learn systems you can trust.
I have talked to VCs a number of times. Always wound up finding other funding in the end, but got a lot of perspective on what VCs are looking for and how they anticipate getting it.
Number one: a VC expects 5x-10x return on investment. That return is typically from selling to another company (which may itself be a VC). It may also be from revenue but VCs these days are less interested in active revenue. They want to sell the company and move on.
Number two: They don't care so much about the actual technology. It's important to have good demos etc, but it's kind of a tertiary concern. VCs are investing, first and foremost, in people who they think will get it done. Not just on the tech side, but on the business side.
Number three: The more you talk about the tech's details, the worse it gets. Focus on why this makes sense to consumers at a high level. Focus on why YOU are special. Not the algorithm. YOU.
Number four: It is enormously helpful to have important industry people to vouch for you. Find filmmakers or producers or somebody who are successful with recognized achievements, and have them write mini-recommendations for the tech.
As engineers, it is very tempting to try and explain to a VC why your algorithm is so much more clever than what's out there. Do not bother. That's not what capital is about.
(1) The team is most important. Have you assembled a team with the necessary experience to convert an idea into a successful product. VCs mostly invest in people, not so much ideas. Ideas are plentiful, people that can ***deliver*** are not.
Parent is dead on. You really need to have an excellent team, and because you're seeking financing that excellent team has to have the resumes to prove their excellence. Find someone who has actually brought one or more products like yours to market and bring them on board. It may cost a lot, but it greatly increases your chances, both of success and of getting investors on-board.
Make sure they sign an NDA before you let them see anything.
They aren't going to buy it from you if they can steal it for free.
while(I produce revenue) {
give me money ;
}
take most of whatever is left ;
buy yourself something nice ;
First patent your algorithm
Then license it to chip makers, electronics companies, apple, samsung, etc
Find a sales guy with good contacts and let him sell
Almost every major product these days licenses thousands of patents from tiny start ups like yours for different things
The BBC's Dragon's Den has some eye-opening examples of a broad range of pitches, and is actually quite revealing of the difference in mindset between investor and investor.
Avoid powerpoint altogether. Turn up with a cheap video camera, let the investors shoot video themselves, process it on the spot and show an A/B comparison. Then be extremely prepared to answer questions, pulling up slides and more demos on demand as necessary. It's about business model, not tech. Given YouTube (as just one example) is adding more impressive post-processing options for free - their Calibration-Free Rolling Shutter Removal is pretty cool - what does your stuff do that is so different? How is your product positioned in the market, who are the other major competitors, how do they make money, what is the overall market size for this segment of tools, what is your marketing plan, how will you fend off competitors, who are the business partners, etc etc.
Far better to delay the VC route if you can. Find a strategic, patient, trustworthy and experienced angel partner. VC, from what I've seen, attempt to steer companies to where they perceive the market to be, and this is likely very different from own ideas.
but how fast can you reach stratospheric levels. They want to be at the bottom of a very step revenue (and profit) curve so the can quickly make many time stehir investment.
I'm a consultant - I convert gibberish into cash-flow.
I specialize in algorithms and have worked as a principal in VC-backed companies for over 15 years.
They won't care about fine details but they have to understand enough to know:
1) How easy is it to replicate your algorithm?
2) What enhancements/future features do you have in mind to stay ahead of the fast followers?
3) Why are you and your team uniquely able to pull this off?
Give enough details -- in fact, I'd suggest slides directly addressing these questions -- so you can back up the answers.
As others have said, 'It is all about how the business makes cash.'
Who buys it, why and for how much money?
Is it an online processing via a web portal & FTP or does the customer license for use on a PC/Mac/Linux?
Bo.
From art of the start by guy kawasaki: 10 slides, 20 minutes, 30 point font. If you haven't read the book, you should, it's an easy read.
sysadmins and parents of newborns get the same amount of sleep.
Wow I can take my old crappy video tapes digitize them and make them look good. I would love to do that and would pay for the service or software to do that. Do a kickstarter wrap you program up in an easy to use interface and you have a product.
A list of patentable features of the algorithm would be right at the top of my list as well. If you can also show lack of existing patents covering any aspects of your code that would be good too.
"Malo periculosam, libertatem quam quietam servitutem." -- Jefferson
This is currently modded funny, but that might actually work. Even if it's just an interim step to generate a little income.
Plan My Week for iPhone
This is the hard part. You have created something that you are proud of and you want to share it with everyone what you did and how hard it was. The hard truth is no one gives a shit. No one cares how many long nights you spend pounding out code, what clever math trick you discovered and created. No one cares about how many total man hours went into your project or even how the thing works.
The bottom line is what your product does and how well does it do it. When your car breaks down, all you care about is what is broken and how much to fix it. You don't care how much work is involved in replacing the broken part, what other parts that needed to be removed to get at it. Details like the broken part has a funny unique shaped nut that required a special socket that was only on your model car. Yes, those details are important to the mechanic but you still don't care. You just want a working car.
Same as the algorithm. No one will care how it works. It only matters that it does work. You only want to include a brief, brief overview to explain what it does and why it is different than anything else that may exist. Beyond that, you need a business plan. How your product is going to be sold. Best case scenario, worst case scenario. Future growth, for both the market and your product. Is this it for your algorithm? Is this the best and only thing you can sell or will a 2.0 version be possible?
Venture Capitalists are brainless, visionless bankers. They will invest money if it looks/behaves like something else that has already made money.
Business plans? they are a dime a dozen. There are companies that will produce an incredible business plan for you, and with the better business plan companies, you won't even have to answer too many questions - they will do the research and produce the projections for you. And it will be "VC Ready".
If I was in your shoes, I wouldn't even try to sell this to VC's. I would first create and expose an API for content companies to pass their clips through, charging a few cents per 100 frames, and returning to them an improved version of their clips. Now suddenly you have a product, and you can start offering this product to web sites that accept clips from users. Score one or two such customers, and your business plan people have incredible stuff to work with to make your business plans look extremely appetizing.
And if you decide you do want to do that API thing and turn your technology into a product, I can help you guys. Just contact me here :-)
All those moments will be lost in time, like tears in rain... time... to... die...
And clips from CSI
Venture capitalists are not interested in algorithms, they are interested in a business plan they can believe it. Come up with a way to make money using the algorithm and "sell" it too them.
Don't go the VC route: there is enough slashdot anecdotal evidence to make it clear that being impoverished, enslaved, and destroyed is a repeatable experiment.
Don't take it to Apple: The Stoltzfus' (who invented Rosetta Stone) tried that with a previous product, and their failure and loss -- and the point that you never heard of the previous loss are an object lesson.
My advice? Contact the Stoltzfus' in Harrisonburg Va, and ask them to give advice (as a mentor). They might agree. Then give your basic presentation to them. Then listen to the advice, and market the idea yourselves.
Correct Horse Battery Staple: 72 bits of entropy. Enter "Correct H" into google. When it generates the phrase, that's
Private equity people who have industry insider knowledge about digital video or a company that wants to partner for being able to co-brand or use the technology may be a better partner where they can see a benefit to their current work, clients or product output.
These people may value the product and investment higher than a VC...
I have gone through this before. I have had a pitch where 100+ VCs would not return with an email or other response.
Get the right insider who used to be with "the competition" or "the industry leader" or "just sold his company" in this field and you can see a deal in a few weeks. Why? Because they already know the market, difficulties and reason why your answer is what is needed for people's needs. These types of people have their own sources and often sell themselves on the deal. That is what you want to aim for.
Plus you are likely to see a higher valuation on your work. Instead of you and your partner getting 20% of the company, maybe you both split 40% or more.
When you are pitching to venture capitalists, it will be the rare exception where you can expect any sort of confidentiality. VCs will act in their own interests, and if they think that leaking/sharing something you told them will help make them money, they will do it. They don't care about the details of your algorithms. They probably won't understand what you say anyway. They only care about what your technology will do for them (i.e. make them piles money). More important than any technical details is the assurance that you can protect yourself from copycats. And of course they also want some assurance that you can be trusted to make money with their money, should they give it to you. Use your PowerPoint deck to tell them about your past successes, your credentials, your well researched business plan, how you are uniquely qualified to make their investors money, and how much interest you've been getting from competing sources of funding.
The people who would most benefit from this technology just don't care about quality. They are the ones who shoot video on their crappy phone handset and upload it at a terrible resolution onto YouTube. Oddly, this has led to the situation where my TV is now capable of playing back full HD content but mostly ends up displaying low grade badly shot footage that looks like it come from 1990's Russia. We can thank YouTube, camera phones, and Tosh.0 for much of this. News broadcasts which I pick up in HD 720/1080p often look like a blurry mess until the feed switches back to the studio. Feeds from security cameras on the street and in shops are even worse. How they expect you to be able to identify a suspect from a 4fps 160x120 grainy picture is beyond me.
At the other end are the people who care about quality and might like your product. They are most likely using high quality consumer / prosumer / pro grade cameras, good lighting, and a tripod. They will receive a much reduced benefit from anything you have developed.
I guess your target market then is YouTube and the other video services. They add it to their back end processes and apply it to any uploads as needed.
All those moments will be lost in time, like tears in rain.
How often do you think these are these hypothetical questions that someone posts on Slashdot so they can use the great ideas here to write a magazine article or book? There are quite a lot of good answers here...
I recommend you take this question to news.ycominator.com It is a forum were lots of entrepreneurs meet and discuss. Most likely, in Slanshot you will only get responses undervaluing your job/algorithm, saying how many ways it sucks and how you could achieve the same with current X or Y (including virtualdub) open source technologies.
Good luck!
Ubuntu is an African word meaning 'I can't configure Debian'
From the NY Times last Thursday, a fun article about VCs. It won't answer all (or maybe any) of your questions but it might give you a sense of what you're thinking of getting into.
Dear Slashdot: next time you want to mess with the site, add a rich-text editor for comments.
I suppose, this sounds like a fully-automated video normalization - instagram for video. If you bootstraped that for 3 years, it wouldn't be hard to write a web app in a couple of months and check that. Apparently, according to Google, previous "instagrams for video" were sold for quite a bit of money (tens of millions) but were not very useful. If yours really works where the other apps fail, that's a win.
See subject.
That's how everyone else seems to be doing it now.
If you have good video enhancement software, don't go strait to a VC. The government (read: military) has plenty of programs which will pay you to develop your idea into a tool which will be used with established systems.
If, for some reason, you don't want the military to get its hands on your work, don't try commercialize it. You may as well get the "free" investment out of it. to
At my work, we recently had a workshop on how technical folks can better present ideas when trying to get funding. Here are a few of the take away points that might help you:
1) Don't focus too much on the technical details.
2) Know your competition and be able to explain why your idea will work better.
3) Practice your pitch beforehand, and solicit feedback on how to make it better.
One recommended method for getting feedback was to grab a small group and ask them to pretend to be investors. Have a few of them focus on what parts of the presentation worked well and shouldn't be changed. Have others focus on what parts of the presentation had problems. When getting the feedback, just take notes but don't try and respond immediately to the feedback. Do this a few times and your pitch will get significantly better. You may end up feeling like a used car salesman at the end of the day. But if you're going to sell your idea, you need to be able to wear that hat for a day.
Have you heard of pagerank? The algorithm which google ran on? Larry and Sergy couldn't sell it. Your algorithm is worth less.
Sell a product. If you don't have the skills (or time) to make a "Photoshop for videos", make a one-touch "video-improver" app. Explain how it could be extended.
Actually, the first question is, why are you talking to VC? You need to raise money, for what? As other mentioned VC are interested in business not so much IP.
VC like proven cash flows, and with them, you can get much better terms from VC, if you have already demonstrated sales, and self funded expansion.
NEWS MEDIA
Your quickest cash is from TV programs which use amateur video, and would be willing to spend significant sums to upgrade video for better broadcast quality. Comedy Central's "Tosh.0" and Warner Bros. syndicated show "TMZ", use immense amounts low quality video. Sell turn key systems to them to generate your quickest cash flow hit. National broadcast news, such as ABC, NBC, CBS, Fox, CNN, MS-NBC, Bloomberg, PBS, and similar, use low quality video on occasion. You want to serve the national head quarters of each, and new media in biggest cities, starting with New York City. There are 52 (cities) US Census Metropolitan Statistical Areas, with more than 1 million people in the US, that would be logical targets for such media organizations sales, which would be logical destinations for turn key solutions if the software and hardware are reliable, and straightforward to use. These same news organizations have offices around the world, such as London, Tokyo, Middle East, and so many other places.
TECH SUPPORT
Sales requires technical support. A web service may be the best emergency back up in the event of problems with turnkey hardware. Online documentation and email tech support during business hours is easiest to do. Tech support that is more complete, such as with call centers is extremely expensive, and would require maintenance contracts, and considerable sales of both software and service contracts.
FEDERAL GOVERNMENT
Federal Government processes large amounts of data, and care about speed and quality. Success with the news market would be helpful evidence that the Federal Government should pay attention. Uncle Sam, like news media, might be willing to pay sharp premiums over what mom and dad consumers would pay. This procurement process is a slow process, and you likely need an expert in Federal procurement, to get GSA vendor number, etc, but if your software is adapted by these organization, you could see major revenue from large numbers of sales at a premium. I would focus on national security market (DoD, CIA, NSA, FBI, and similar) as well as law enforcement, (FBI, DHS, DEA, US Marshals, ATF, US Coast Guard, US state level investigation organizations, and similar). These markets have hundreds of offices, that could use a high quality tool chain to see amateur video, such as of the Boston Marathon 2013 bombing, and similar. The problem is that the software would likely need review for more classified uses, on computers you would not be able to sell to the Federal Government, requiring you to partner with an authorized vendor of hardware, so you can have standardized and tested hardware and software as a "turnkey" solution. Domestic US law enforcement, reviewing police car dash board cameras and similar, would have lower security requirements than most classified US Federal service. Learning what Orange book computers are, and getting security clearances for personnel are extremely expensive, slow, and intrusive, but if the sales are large enough, classified sales could be well worth it. Typically small companies are asked to join forces with very large Federal contractors, to pursue larger and more serious sales, which all takes time, money, and legal fees. The last and most complex hurdle is the US Federal budget cycle, currently under "Sequester", making Federal Sales extremely difficult to achieve.
WEB SERVICE
If your tools are ready to be sold as a web accessible service, you could directly sell to much larger audiences. The web service business model gives you the most control of hardware and software, but you can only sell to customers with appropriate band width. The web service would be credit card, and premium pricing, such as a third to a half the price for a single session as the entire shirk wrapped software package, provided the testing many outputs from the one uploaded video, This woul
"How Do You Sell an Algorithm To Venture Capitalists?"
First of all don't use Powerpoint. And by that I mean DO use Powerpoint and say the word Powerpoint as many times as you can.
The most important thing is to show them that there is a demand for what you do. This is where great ideas make it or fail. Henry Ford didn't invent mass production. It's just that nobody thought there would be enough demand for so many automobiles. Same goes for every other successful enterprises. Who would want a personal computer? Who would want a phone without physical buttons?
So who wants your stuff? How much do they want it? How much are they willing to pay for it?
Drop that sorry excuse of a program and use Keynote. That will easily give you a more professional image. You guys don't have a Mac? Buy a MacBook even if it's only for your presentation. Think of it as an investment for your VC pitch.
Get free satoshi (Bitcoin) and Dogecoins
Not all VCs are alike. If you are pitching to a group, some may be more interested in your business acumen while others will be more interested in your technical ability to stay ahead of the competition. VCs usually have more options with a business with great tech but poor management than one with great management but poor tech.
First, at lest start the patent registration process. Talk about the algorithm enough to convince them you have the technical chops to handle any issues that come up. You should also come prepared with one of your top tech guys to handle any questions that come from the VC's technical advisor. But, mostly focus on the business. Competitive advantages of your algorithm, who is going to buy it, how you are going to market it to them, potential revenue stream etc.
The Startup Owner's Manual, by Steve Blank and Bob Dorf. The startup business has changed. As the book says, "No business plan survives first contact with customers." Their core insight is that a startup is a temporary organization in search of a scalable, repeatable, profitable business model. At the very least, go through the "Business Model Canvas" process described in the book.
The aspects of the technology that you should talk about are those that are directly related to competition, intellectual property, business model and eventually regulations.
Interestingly, those are also the aspects that interest me as a technology entrepreneur.
Software algorithms to improve the quality of images, in particular optics, is already embedded directly inside the latest mobile phones and works in real time. That software takes into account the defects and bias of each type of lenses to produce the best images possible with the cheapest optics possible.
How does your software compare to that? Does it require to know technical details about the camera optics? Is it based on machine learning, and if that is the case, does it need to be trained on particular pictures set for each camera? On what hardware do you already have an implementation, and how fast is it? How fast could it get if optimized?
What is it that makes your algorithm better than the competition? Will it remain better? How much effort would be required for a competitor to reproduce it? Is it patentable?
Is your business model about embedding it in smartphones? If so, are you working with OEMs to integrate it transparently or do you provide it as a separate application? Have you considered a software as a service model?
Could the modifications your algorithm does to the original images cause problems with certain regulations? Some sectors may require to take pictures of videos for study or proof (health and insurance come to mind), would you be able to certify that your technology does not affect a diagnosis? Could this be a barrier to working with OEMs?