A lean canvas is one of the most basic tools of the Lean Startup methodology that was developed to increase success chances for startups.
Typically it takes only a few minutes to fill out a lean canvas, so you should do it the moment a new and exciting idea pops in your mind.
Filling out a lean canvas is the most efficient and productive way to:
There are several versions of the lean canvas, but the Miranor Canvas differs from all the others in the following ways:
Name of your product (service, project, whatever) for which you're filling out this Canvas.
Since for each product you are expected to fill more than one Canvas (one for each MVP), you should add an index number to the name, or better yet some distinct suffix, e.g. "MyProduct - MVPName". The name is just for you convenience, so you'd be able to identify the subject of the Canvas later. You don't have to fill it, but it is recommended that you do, because if you're serious about this, you will probably fill many lean canvases in the near future.
Row I (green) describes your idea and its market in general, large-scale terms.
This row also provides a pool of MVP ideas for the second row. This means that the next rows will probably change drastically as you pivot (iterate on different MVP's), but the first row should remain relatively stable. It's like an anchor that allows you to stay true to your ultimate goal - why you started this project in the first place (besides getting rich, of course).
Row II (cyan) describes the current Minimum Viable Product (MVP) you'd be working on.
Recall that according to the Lean Theory you need to achieve the coveted "Product-Market Fit" as quickly as possible in order to improve success chances. You do this by selecting a specific and limited product (MVP) for a specific and limited audience (MVP market), and test whether they fit each other. If they do - excellent, start growing from there. If not, then move on to the next MVP-market combination. Row II describes this product-market combination, where both components are generally subsets of the general idea outlines in Row I, which is why it is important to fill Row I before Row II.
Row III (violet) outlines the marketing strategy for your MVP (the MVP was outlined in Row II).
After all, a product is useless if no one knows about it, or if no one understands why they should use it.
Describe the problem or pain that you want to solve through this project in one or two short sentences.
Don't go into unessential details or complexities. Use simple words (no techno-talk or professional jargon of any kind) and make it clear enough so that ordinary people could understand the problem just by reading or hearing your description.
This problem description should be (and would be) an excellent opening of your elevator pitch for potential clients, investors, partners, and advisors, so invest some thought into it.
Many people believe that the first step of any startup is to have a brilliant idea, some technological twist or a novelty business model, or something.
But in reality, the first step to any viable, sustainable, and scalable startup is some kind of a real, unresolved pain of a large group of people. Yes, there are some who are willing to buy (or invest in) new technology just for the sake of novelty, but unless you have the marketing skills of Steve Jobs, you'd be very much unlikely to find enough of such people to sustain your business.
On the other hand, most, if not all people are willing to pay for anything that solves their problems. The more pressing, painful, and obvious the problem is, the more they will be willing to pay for a good solution, a simple and effective solution, even if it's not "neat" or "novel", or "cool".
In fact, if you're struggling to find a good startup "idea", try to think of a startup as a solution to a given problem, and then find a problem that is worth solving. Have no fear, there are more than enough problems in the world, sometimes in places you may not have considered. For example, computer games (in fact, any form of entertainment) solve the problem of boredom, which can indeed cause suffering. Get the idea?
If you want to learn more on this subject, you can start with the article "How to Get Startup Ideas" by Paul Graham
Many dog owners lock their dogs in apartments before going to work. The poor animals bark for hours in despair, which leads to quarrels with neighbors, and then everyone suffers.
Who benefits from a solution to the given problem? In other words, who is suffering from the problem? Define at least 3 distinct groups of customers (consumers or businesses) and estimate their numbers.
Now that you have described the problem you want to solve, think of the people or organizations that might benefit from a solution to this problem. Think of what distinguishes each group from all the rest. Use one primary characteristic (two at most) for each group. Each such group is called a market segment.
A characteristic can be a geographical location, living conditions, marital status, socio-economic status, profession, hobby, ownership of some product or property, size of the company (for businesses), and so on.
A characteristic can (and should) also explain why this is a problem for the given segment, if it's not obvious.
No need to define each segment precisely, or specifically, or in detail - that will come later (Row II). Think big, don't restrain your imagination.
If you can't think of enough groups, describe the problem to anyone who'd listen, and ask them who else might have the given problem. This is one of the fields in the Canvas that can benefit from collective wisdom and brainstorming.
Specify the total worldwide number of clients in each segment. No need to be precise, just the order of magnitude, e.g. 10M (ten million), 100M (hundred million), etc. In many cases a simple Google search can help estimate these numbers. You will need them later on.
If your segment consists of individual people (consumers), then you're making a B2C (Business-to-Consumer) solution.
If your segment consists of commercial companies, then you're making a B2B (Business-to-Business) solution.
It is okay to mix B2B and B2C in this box at this stage, but ultimately you will probably have to choose one - mixing these categories or even switching between them at a later date can be difficult, because they often require different marketing methods, skills, and resources. For example, B2C usually means much more customers that pay much less, so the payment system may have to be completely different.
- Neighbors of people with barking dogs (100M)
- Dog owners who get complaints about their dogs (10M)
- Municipalities (1M)
- Animal rights organizations (1K)
How can you generate money from solving the problem? List at least 3 different generators.
A revenue generator is the basis of a business model for your startup. It usually consists of a potential solution (product or service) and a preferred revenue stream for the solution.
A revenue stream is often solution-dependent, which is why each stream should be combined with a solution. However, a stream may also be so generic or self-explanatory that it may stand on its own, without a concrete solution, e.g. "Workshops, seminars" (obviously, the solution is to teach people how to solve the problem for a one-time fee).
Even if your project is not for profit, you must have a sustainable source of income to build and maintain a viable solution. Nothing works for free, at least not for very long. Even Wikipedia, which is a fully volunteer-based project, still needs a lot of cash for its operations.
Fortunately, even for non-profits there are many possibilities for a valid revenue stream, including:
For a commercial project, the possibilities are much more plentiful, of course. In addition to the above sources (yes, commercial project can accept donations and sponsorships, why not), the most common ones include:
- Resell helpful products/services by commission
- Dog daycare centers by subscription
- Dog walker/sitter platform by commission
- Conflict resolution service for a fee
- Crowd-funding
Calculate the total addressable market size, i.e. how much money you could theoretically make each year. Write down the end result and, preferably, briefly explain how you got this result.
The Total Addressable Market (TAM) is one of the most important parts of the Canvas, because its value determines whether the project is worth your efforts at all (sadly, it may well not be).
The TAM is calculated based on the "Segments" and "Revenue generators" boxes that you filled previously.
For each revenue generator you listed, take the total worldwide number of customers that are relevant to this generator (see the "Segments" box for that) and multiply it by the amount each such customer would pay per year (aka ARPA - Average Revenue Per Account).
To estimate the latter value (the ARPA), the best bet is to find a few of these potential customers and asking them how much they are willing to pay per year to solve their problem, regardless of how that problem is solved (say, you have a magic wand that can solve it - seriously, say it). If you can't find potential customers to ask them, then you are probably in big trouble anyway. I mean, you will have to find them when you have a working solution (which should be very soon, according to Lean Theory), or you will fail for sure.
That being said, if for some reason you prefer not to look for potential customers at this point, then here's what you can do instead: imagine yourself being in this customer's place, having the same problem (many entrepreneurs try to solve their own problems, so this part should be easy). Imagine it well, and then ask yourself how much you are willing to pay (per year) to have your problem magically solved. No need to come up with exact figures - a crude, order of magnitude approximation should suffice.
Actually, there is an even simpler method of estimating the ARPA - a five-animal model by Christoph Janz. To summarize it briefly, if your customers are "flies" (low-end consumers) then set ARPA to $10; for "mice" (high-end consumers or "pro-sumers") - $100, for "rabbits" (developers, designers, small businesses, or very rich consumers) - $1,000; for "deers" (small to medium size companies) - $10,000; and for "elephants" (big companies and enterprises) - $100,000.
Once you've calculated the market size for each revenue generator, pick the largest one and that will be your TAM.
Note that you may be tempted to sum up all the sizes to increase your TAM. Don't - it's usually pointless. We're talking about order of magnitude approximations, not exact figures. Often, one of the generators is order(s) of magnitude larger than the other, in which case summing them up changes nothing. And if two generators are of the same order of magnitude, then their sum will still be of the same order of magnitude. Besides, different generators often have overlapping customer bases, and you can't expect the same customer to pay twice for solving the same problem, so the sum would be meaningless as well as pointless.
The TAM should be at least $1B (one billion US dollars per year worldwide). This may seem like a lot at first, but once you account for conversion rate (how many potential clients become actual paying customers), competition (what percent of the market you can possibly take), and subtract operational expenses, the remaining figure will look much less impressive, which is why it is important to have a very large figure to begin with.
In fact, most investors will probably lose interest in you the moment they find out your TAM is less than a billion dollars. Therefore, if you got a figure smaller than $1B (at that moment you will be tempted to sum up different generators to pull it up to $1B, but do try to resist), then it may not be worth your while to continue with this project. Of course, you can always return to the previous boxes and come up with other revenue generators and segments in an effort to find a more potent combination.
By the way, if you're still having trouble believing that $1B is a small market, then consider the fact that the TAM for dog food in US alone is $40B, and it's still a medium-sized market! What is then a big market, you ask? Well, in 2014 the worldwide TAM of crude oil was about two trillion US dollars. The global e-commerce market is expected to reach $2T as well in 2015. Now that's a number that gets investors REALLY interested. On the other hand, the number you write down in this box should be realistic, and not ridiculously large.
If the market size is less than $1B, but has the potential to grow rapidly in the next few years, then it's worth pursuing. In fact, it's even better, because you'll be in a position to monopolize the market once it grows.
This most often happens when your product relies on a new technology that is essential to the solution (i.e. solving the problem without it is much less practical), but not prolific yet, e.g. latest phones, newest browsers, smart-watches, 3D printers, lab-on-a-chip, gigabit networks, quantum communication, etc.
Of course, such approach also carries great risk - what if you're wrong and the market won't expand? Satellite phones, mini-CDs, netbooks, 3D TV, online coupons... so many promising innovations that didn't live up to the expectations.
- 100M neighbors * $100/year = $10B/year
- 10M dog owners * $100/year = $1B/year
- 1M municipalities * $10K = $10B/year
-----
TAM = $10B/year
Describe a minimalistic solution to the problem. A few main features, no technical details.
A solution is the basis for the MVP (Minimal Viable Product), which is the simplest possible product/service that can attract customers. It's not a complete solution to the problem, but it does represent a first step toward the ultimate solution, and provides some kind of value to at least some of the clients that have this problem (see the "Segments" box). But at the same time it has to be a solution so simple, it could be deployed within a very short time, perhaps days, weeks at most.
To sum it up, you need to think of a partial, temporary solution that is very easy and quick to build, but that does solve something related to the given problem at least for some potential clients. A partial solution can be something that just lessens the pain, or perhaps merely promises to solve the problem, nothing more. If you don't believe a mere promise can be good enough, think of pre-ordering or crowd-funding, where people pay real money for a solution that doesn't exist yet.
If you have trouble coming up with a solution that meets the above criteria, then try the following:
At this point your top priority is speed - the essence of Lean Startup. It is very important to not only remove non-vital features that will slow you down, but also to find the cheapest and fastest way of implementing the remaining ones. The last thing on your mind at this point should be whether the solution is scalable. Will it support millions of users? No one cares, and neither should you. The point of MVP is to run a quick experiment to see whether there is enough interest in your ultimate product. If there is, then, and only then, you start thinking about growth, about making something that scales.
That being said, there are a few features that are essential for any MVP, especially an app or a web service, even though they don't provide much value for the customers:
Online resource/forum for sharing pain, and endorsing existing solutions: dog trainers, dog sitters, dog walkers, daycare centers, training devices, etc.
Who exactly are the potential customers for the MVP described in the "Solution" box?
Now that you know what your solution is, you should also see the appropriate audience for it - the so called Early Adopters.
Look at the "Segments" box (right above the "Segment"), and select one segment that fits the solution the most (i.e. gets most benefit), in your opinion.
Now, try to narrow down this segment by adding constraints on it (geographical location, age, status, etc). The "Solution" box should help you with this.
It is important for the initial segment to be very narrow and specific, even if it makes it too small to be of any commercial interest, because as a rule the more specific the audience is, the easier it will be to get to them.
Don't worry, once you find the product-market fit, you can start widening the segment, but no sense doing it at this point.
People who suffer from barking/howling neighborhood dogs in the local area.
How exactly will you make money from your MVP (if at all)?
Look again at the "Revenue generators" box (right above "Revenue stream"). Recall that a generator is generally a combination of a solution and a revenue stream.
So, take the generator on which your current solution is based (see the "Solution" box), and write down the revenue stream that is associated with it, only this time be more specific. You can even go as far as setting a price tag.
Here is a quick recap of the most popular revenue streams:
Note however, that since an MVP is all about speed, some of the above streams may not be effective at this point.
For example, selling statistics requires a large existing user base before you can assemble enough information to attract customers (whoever is willing to pay for this information), and that will take too long - too risky at this stage.
Another example is government funding - even if you're eligible for it, it usually takes a long time to get through all the beurocracy - again, too risky.
The best streams are the ones where your users are also your customers (the ones that pay). This is called a 1-tier business, and it is usually by far the easiest one to build.
It is also possible to have a goal other than revenues at this stage.
For example, you may be trying to increase awareness of the problem you seek to solve, or pursue some other, noble, non-financial goal.
If that is the case, then instead of a revenue stream write down your goal in this box.
However, don't be tempted to do so only because taking money for a partial solution feels uncomfortable for you. You should test your business model as soon as possible, or you risk wasting too much time building an unsustainable product, so use this "exempt" only if there is no other practical option.
PPC and/or commissions from product/service providers referred and/or recommended by the site. However, the main goal is to increase awareness.
What parameter(s) will you measure to determine whether there is a product-market fit?
Recall that the main purpose of an MVP is to run an experiment that measures the product-market fit.
How do we measure it? By pre-selecting one (rarely more) easily measurable parameter. The trick is to a) choose the right parameter, and b) do so in advance.
Why in advance? Because otherwise it is all too easy to fall into a trap known as post-hoc fallacy, when you redefine the experiment after-the-fact to get the results you desire. This is why the convenient "let's just do it and see what happens" stratagem is risky - it may create a false sense of success until one day you suddenly find yourself bankrupt, or worse yet, waste years on a "zombie project" that isn't going anywhere.
Another reason for choosing a metric in advance is so that you could actually measure it. Not every metric can be measured trivially - some require you to install a more complex analytic, perhaps even a customized, built-in measurement mechanism. This is why the metric belongs to the MVP row of the Canvas - it defines the final feature of the MVP, the proper measurement system.
But even if you choose the metric in advance, it still has to be a good one - the one that can truly measure success, otherwise all this would be pointless.
The selection of a good metric depends greatly on the revenue stream (the previous box).
Revenue stream | Suggested metric |
---|---|
Pay per unit | Number of units sold per month |
Paid monthly subscription | Number of subscribers per month |
Advertisement (website/app) | Monthly active users |
Freemium (free+premium) | Number of premium subscribers per month and conversion rate from free to premium |
None (the goal is to create awareness) | Total unique visitors that spend more than 30 seconds on site |
Of course, these are only suggestions, but the underlying principle is to choose the metric that measures the relevant success, and not just something that only looks impressive - the so called vanity metrics.
Note that the vanity metrics are important too - they boost the company's image and self-esteem, create brand value, impress visitors and (some) investors, and more. However, they don't necessarily provide a reliable indication of the product-market fit, so they're usually not relevant in this context, and should be ignored (or rather measured and put away for later).
Some schools of thought recommend adding time and quantity constraints on the metric. In other words, you should not only select a proper metric, but to specify in advance the minimal result that you want to get in a specified time frame in order to declare victory. For example, "100 in-app purchases within the first week" or "1000 e-mail subscribers within 30 days".
However, such specific details usually require a deep understanding and experience, as well as knowing your market extremely well, which is probably not true.
Fortunately, it doesn't matter much. When the product-market fit occurs, you will see it (as long as you're looking at the right metric, of course) - the numbers will start to increase exponentially, i.e. by a factor of 'x' every fixed time interval. For example, 100 engagements on the first week, 150 on the second, 225 on the third, 338 on the fourth and so on (i.e. each week 1.5 times more than the week before). Of course, the numbers don't have to be so precise, but you get the general idea. By the way, 1.5 is reportedly the most realistic factor across many fields.
That being said, you should set some time limit on the experiment. It can be flexible, but not too much, otherwise you're still risking it with a zombie project - waiting too long for the exponential growth to begin.
The time limit depends greatly on the product, but for apps and websites the recommended time limit is generally no more than 2-4 weeks. If you don't see enough growth after that time, something is wrong - better return here and pivot - try a different MVP.
- Number of clicks on the referred/recommended solutions
- Number of non-profitable engagements (shares and endorsements) to measure awareness.
Where exactly will you find your target audience? List 3-4 different channels.
Now that the MVP has been fully specified, it's time to think how you're going to market it.
The first step is to find a place (physical or virtual) where your potential customers can be found easily and efficiently. Of course, you have to have access to this place - the ability (and permission) to communicate with the people in it. Such a place, together with the method of communication is the marketing channel you seek.
In general, the more channels you have, the more customers you are likely to find. On the other hand, each channel requires much individual attention, and maintaining too many channels in parallel can be exhausting and ineffective. In most cases, 3-4 channels seems like an optimal compromise between these two constraints.
Now, where do you find a channel? Well, you have already defined your early adopters in the "Segment" section above. The more you know about these adopters, the easier it will be for you to find a proper channel.
The main reason why the above channels aren't so good is because they're unfocused - social networks and search engines engage large and versatile audiences, only a tiny percentage of whom are your early adopters, and you usually have direct access only to a tiny fraction of the network. Don't be fooled by the idea that these channels are free, and so it doesn't matter. They are not free - you are paying for them with your time and effort. The less focused the promotion is the more effort it requires, until you're just not sure whether the campaign isn't working because there is no product-market fit, or because you can't get to it - the effort required to sift through a large audience is simply beyond your abilities. The best, if not the only way to leverage such a channel is to achieve a virality, where your own customers help you with your campaign, and an average customer brings you more than one new customer. Unfortunately, virality is very difficult to achieve - it requires very specific conditions and a fair amount of luck, so unless you have a repeated experience in creating viral campaigns in a similar field, you shouldn't rely on it to happen.
These are the paid advertisement platforms, starting with Google AdWords and Facebook campaigns. These are good for growth that comes after the product-market fit and validation of a business model, but not to find early adopters.
First of all, these channels cost some real money, not just for views or clicks themselves, but in most cases also for creating professional-grade design for your campaign.
But a more troubling issue is that by turning to the paid advertising you're basically give the controls to a machine. As sophisticated is it might be, it doesn't know your project or understand your audience, so the efficiency of such a campaign will be very small in most cases, and again, if it fails you cannot know whether it failed because the product is wrong, or because the machine was looking in the wrong place. Still, it may be better than nothing, but then again, if you don't know where to find your audience on your own, then perhaps you should rethink the whole thing. In any case, note that if you mention to a potential investor that your primary channel is paid advertising, it will probably reflect poorly on you.
- Animal rights organisations
- User-based awareness campaigns
- Paper leaflets distributed by early adopters
- Facebook referrals
List at least 6 alternative solutions to the same problem that already exist.
In this context, we're talking about competition in a broad sense. Not just direct competitors like other companies in the same market, but the alternative ways of solving the same problem, perhaps not with marketed products at all.
In fact, in most cases the first and foremost item on the list should be not solving the problem at all. A lack of solution can also be an alternative.
To illustrate this, let's look at, say, the Toyota company. What is its competition? Other car manufacturing companies? Definitely. But that's just the tip of the iceberg. Let's take a broader look. What about other forms of transportation? Planes, ships, trains, bicycles even - all these are alternatives to all cars, including Toyota. And it goes even further. What about walking? Yes, our legs also compete with the car companies. But this means that the list of competitors has to include environmental organizations and sports clubs that promote walking as an alternative to cars. But it doesn't stop even there. Let's think about it even more generally. Why do we need cars at all? Well, for example, so we can drive to another city on a weekend to visit our relatives. But what if instead of visiting them we could just have a video conference with them? Not the same thing, of course, but beats a 4-hour drive. Who would've thunk it - Skype is an indirect competitor to Toyota. Get the idea?
The list of competitors will become useful in the next box, but its main purpose is for you to assess how well you know the market and the problem you're dealing with. If you cannot name enough alternatives, then something is wrong.
Never say you don't have any competition. It irritates investors, and makes other entrepreneurs ridicule you.
If you still believe there is no competition, then forget this word, and instead think of alternatives, as described above. Still nothing? Then think of it this way - if there is no alternative to your product, then how do people get by without it right now?
- Silent suffering
- Looking for solutions and support on your own on Google
- Wining and/or asking for solutions on Facebook
- Looking for solutions and support on your own
- Complaining to police or municipal services
- Promotional platforms for products that may solve the problem (Ebay, AdWords, etc)
Why should anyone use your product? What value do they get from it? List 3 such values.
A Unique Value Proposition (UVP) is a promise you make to your potential customers. It is the basis for the marketing message of your marketing campaign, so give it some extra thought.
Or just follow this algorithm:
This is one of the things novice entrepreneurs have most trouble with - distinguishing between a feature and a value. Without understanding this difference, the "Unique Value" box looks too much like the "Solution" box, and the marketing message looks more like a technical spec, a list of features.
To reiterate, a feature is something the product does, and a value is what the product is for.
No, it's not the same, because most customers are interested in a value, but not in a feature. Even though values are directly derived from features, this derivation is usually not trivial, and must be made in advance by you - most customers won't do it on their own.
Emotional value: provides solace/stability/support/sense of control/hope, reduces frustration/stress/anxiety, entertains, etc.
Rational value: saves time/money, educates, optimizes work flow, grows business, protects nature (yes, rational doesn't mean greedy), etc.
- You're not alone (emotional and social support)
- Control the problem, don't let it control you
- Hope for a noise-free future
What stops someone who sees your product or service from copying it and grow faster than you?
Unfair advantage, aka "secret sauce", is some factor that protects you from direct competitors, as outlined in the "Competition" box.
It is not exactly part of your marketing strategy to customers (that would be the "Unique Value"), but it is an important piece of "marketing" to partners, investors, and even other team members. If you don't have any advantage over the competition, then your business is too risky, so you should definitely have an advantage, and a good one too - the one that is very difficult to replicate.
These examples are not as good as the previous ones, because they are relatively easy to replicate. Once you achieve some minimal success, you'll "appear on the radar", and someone will try to replicate your success. If you don't have a secret ingredient, then it's mostly a matter of who has more resources. The potential competitor can raise money, get access to your not-so-secret-anymore market, and hire a highly professional team that will kick your butt off the map.
Still, these things do provide protection at least from amateur potential competitors with little or no resources, so it's still better than nothing.
"Our advantage is the fact that no one else has thought of our idea."
This one is so bad that you should never say it, not even think it. Most experienced entrepreneurs these days will laugh at you for saying such a thing, and most investors will scoff.
The fact is that if you thought of this idea, then a thousand people all over the world thought of it too, and some of them are already working on it (and some may have already succeeded, or worse, failed - you just haven't found them yet). And even if for some reason you are the first to come up with the idea, the moment you go to the market, it will become known potentially to anyone, unless your plan is to stay in the shadows until the exit, but then how will you market your product?
The only conceivable exception to this rule is when you are a top expert in the field, and your knowledge is so deep and cutting-edge, that only a few people in the world could think of this idea, and you know for a fact that they didn't because you know them all. But if that is the case, then you already have a good advantage (the first one in the above list).
The most important conclusion from the above passage is that you shouldn't fear talking about your idea, and in particular to fill and share this canvas. If you have an unfair advantage, then it won't help your competitors even if you tell them what it is. And if you don't have an advantage, then keeping secrets will make no difference.
In any case, of all the people who find out about your idea, 99% can't steal it because they don't know how. And of those few that do know, 99% are already too busy working on their own, much better ideas (in their eyes). And of those extreme few that just happen to be free right now AND want to implement your idea AND actually can do it... well, you just found yourself a co-founder.
In other words, the chances of someone stealing your idea (not to mention succeeding with it) are ridiculously small, and at the same time the benefits of sharing your idea and getting a valuable feedback and even more valuable exposure are real.
Don't get discouraged if you cannot fill this box with anything good. If you don't have an unfair advantage, then you can settle for a fair one - simply being better, faster, more creative than your competitors. It's very hard, and it probably won't score you much points with potential investors, but it is doable and commendable.
I have 3 neighbors with their dogs howling all day long, every freaking day, and it drives me crazy! So, I have a deep understanding of the problem and an extreme personal motivation to solving it.
Need help with the canvas? Found a problem? Have a suggestion? Contact Us
© Copyright 2015 by Miranor. All rights reserved. By using this site you agree to the Terms of Use and Privacy Policy.
Page last updated on May 12, 2019.