This view is (hopefully!) from an investor's perspective, and treats 'traction' as different from internal management metrics.

What is it?

I've always liked Naval Ravikant's basic definition: 'Quantitative evidence of market demand.'

Traction is proof that somebody wants your product. Ideally, it should communicate momentum in market adoption.

What works?

Broadly, I find traction most convincing in the following order:
- Profitability
- Revenues
- Active users
- Registered users
- Engagement
- Partnerships/clients
- Traffic

So that's the broad version, which only half applies to any startup, as all are different. Here's where it gets interesting: use evidence creatively (and honestly) to tell a unique story of your  momentum.

How to do it: some examples

Use specific numbers (ex: 50,000 registered users) or growth numbers (ex: 30% monthly growth in users), to best communicate momentum. To really push it, include a graph of your accelerating growth (legend has it that Google's pitch deck was one slide: their user growth).

Let's say it's an e-commerce startup. You may want to talk about revenue growth, average sales or gross margins, customers, average return visits per customer, units moved, etc., as well as some industry specific numbers I'm probably forgetting. And think about the growth.

Or a fashion Q&A consumer internet product (there have been a few cool ones recently). You may want to talk about pure user numbers, engagement of those users (number of questions or answers, photo uploads, time on site, outfits tried), virality (average new invites per user, waiting lists, etc.) or partnerships (brands signed on, stores engaged, etc.).

Say a freemium SaaS product for small businesses. You may want to talk about revenues, conversion to paid customers, registered users, cost of customer acquisition + lifetime customer value, signed distribution partners, or if you're really creative, some quantifiable evidence of value to customers. And always consider using growth numbers.

FInally, one more common type: Enterprise. What can work here is revenues, number of enterprise customers/clients, average contract size (growing?), and your qualified sales pipeline. Try to find some downstream value add example, say that your product engages X of their staff or customers, or saves them Y % on marketing costs while improving results. An example of customer progress can also work well ('started with limited pilot, moved to full rollout across Z offices').

Try to be imaginative about using the evidence that works for your business, while still including the basics that investors will want to see. Always make sure what you use shows momentum, and is honest.

Engagement: I think of traction in terms of engagement by your target stakeholder groups. If you are building a development platform, look at the number and quality of developers using it. If you are building a consumer play, how many consumers are engaging with your product?

If this is all new to you, I would recommend reading Dave McClure's AARRR methodology as a starting point, and then customize for what is appropriate for you. See: http://500hats.typepad.com/500bl...

Revenue: I think revenue is super important, but the result of a long list of precursor actions. Most of the startups I work with are doing great with the precursors, but haven't actually made a lot of revenue yet. That is totally fine, as long as they are getting the initial engagement.

Cost Per Action/Acquisition: I think measuring acquisition cost is absolutely crucial for any startup, but completely useless without a measure of Lifetime Customer Value. The game is simple, the cost to get a customer (CPA) needs to be lower than what you make off of your customer (LCV). I fully expect CPA to go down while LCV to increase over the life of a startup, so I consider engagement measures (do I have a product anyone cares about?) to be much more important in the beginning.

Press: I simply view the press as a tool to get acquisition cost down.

I don't think profitability is a immediate and convincing sign of traction in a start up. Sure, profitability is a good barometer of how well a company is performing financially but it's not a good indication of market penetration or product/service uptake/usage.

If you had 1 customer after 6mths in operation paying you $10k/mth with a cost base of $500/mth then you are profitable but you certainly don't have traction. You only have 1 customer, which is obviously a risk, lose that 1 customer and you fail.

Twitter is a great example of this, do they have traction? - YES , Are they profitable? - NO

These would be my priorities

1. Engagement
2. Traffic
3.1. Users/Clients
3.2. Partnerships
4. Revenues
5. Profitability
- Profitability can only be achieved when you have revenue- Revenue can only be achieved once you have paying clients (I don't see investment capital as revenue)- Clients are a direct result of engagement and traffic.- Acceleration of client acquisition can be found via partnerships
Engagement is critical. I have seen many great startups fail because they don't or can't engage with their target market. Without engagement you can't even begin to measure traction as you have none.

As Geoff and Bart said already, it is about proof of market value for this specific startup. It can be a lot of things but basically it requires outsiders to deal with the product/service the startup proposes.
However, one missing point in the previous comments is the dynamic of traction. Not only do you need to have good metrics but they must also trend north. Flat figures do not make it. Momentum is key to define traction.

We define it by customer or revenue traction. To be fair, depending on your experience and reputation as an entrepreneur, the bar can move. What might be a set of beta clients for a very experienced entrepreneur might mean 1M in revenue for a first time one.

It can also mean non revenue issues such as team build out, technology completion, partnerships, etc

1) Established Brand - visibility in your vertical market(s) and customers, evangelists, partners, investors "grok" what your company and products/services are about.

2) Revenue Stream(s) in place.

3) Referenceable customers.

4) Web Site Traffic.

5) Social Media Visibility & Acknowledgment by Influencers.

6) High Growth may or may not be good - too much can kill a startup as fast as no customers or traction. Every new startup is not going to be like Google, Groupon, Facebook or dare I say Quora.

7) Localization or Strategic partners outside of US who support and believe in your product, services and company.

8) Cohesive management team that has worked together through rough patches in the road.

9) Sufficient revenue, customers or "visibility" to drive investment (if needed).

10) Customer advocates who evangelize your brand.

In no particular order:

1. Am I getting the right customers, do they see the price-value equation?
2. Can I scale customer acquisition with minimal capex and progressively lower CPA?
3. Can I retain these customers for XX months? Do I know my time to break-even on a customer?
4. Will they recommend my product to the peer group?
5. When I pitch to prospective partners / employees, do they "get it"?
6. Do I have a 30-second pitch on what I do and how this is different from other comparable solutions?

I agree with the above answers and especially with Marc's.

Traction can be defined into stages: proof of concept - people nod their heads at the problem you are tackling; proof of principle - a few early adopters are buying into your solution (this could be customers or end users, depending on the nature of your product or service); market validation - a growing awareness of your solution and finally - market adoption - your solution is now pervasively being applied, copycats are emerging and your email inbox is filling up.

The rate at which traction is occurring, as mentioned by Marc, its velocity is a key determining factor in whether you will get through the various stages before competitors enter the market and whether you will be able to attract investors and ultimately a buyer for your business.

Venture capitalists have a well-trained nose for sniffing out companies that are achieving optimal traction velocity. Ideally, they look for hockey stick velocity or the propensity to achieve this.

The term is typically used to describe a start-up that has customers or, more likely, paying customers.  It means the dogs are eating the dog food; that the company has something that people need and want as evidenced by their buying behavior.  "Traction" gives the investor confidence that the company is on to something and "traction" therefore increases your odds of being funded.

In one sentence: The emerging visibility of patterns of repeatable, scaling usage

A long-dormant thread, but still an active question in the startup community. As the wildly different answers to this question indicate, there's still a lot of confusion and disagreement. Here's my definition:

Traction is the accumulation over time of commitments from paying customers and/or monetizable users obtained from the public marketplace as a result of conscious, repeatable and quantifiable methods or techniques.

Each of those pieces needs some breakdown, which I go into here: Defining Traction, but it squares nicely with Brendan's answer.

This is a function of the type of company.  A pure consumer play may get credit for traction based on its count of engaged users and the trends associated with that count.  A B2B play is more likely to be expected to have paying customers.
 
I think too that there are regional differences in the definition of traction. Investors in areas like SF and NYC are more likely to place bets on concepts with the right teams, technology and trends whether the revenue is yet there or not.  I think the rest of the country, particular the South, is more conservative and has more of a "show me the money" frame of mind.

I agree with folks who see traction as traffic, adoption, and as a sign of engagement by users. I'd only add that in social products and tools, traction includes engagement of users with other users as well as engagement with the product. It's easy for us who design and build these experiences to see traction as confirmation of our product vision. But engagement of users with one another means that the product vision not only translates in user experience terms, but translates in social interactions also.

So I look for communication, for commitment and habitual use, for users adopting product use in different ways and for different purposes, and for users contributing and following based on social signals and perceptions. Those all tell me that the system's architecture facilitates social practices, and that the culture or community of users now adopting the product see reasons for product use that express or manifest their own deeper motives for participating.

How much you spend to get to a certain amount of revenue and users.

Example: spend $0, just hard work and long nights and creativity, to get to $100,200,300,500k or $1MM-10MM.

That's how I define traction.

I want to stay anonymous for now.

My start-up just launched out of public beta and we've signed up 100 paying customers in our first 30 days at price points between $30 - $100 a month.

Would you consider this good traction?

Is the start up on the right path? Simply put, are profits on the horizon? Profits can be in many forms: users, revenue, unique visitors, downloads etc... Its like saying, the baby has taken the first step, so it should start walking and running soon :)