Tag Archives: search engine optimization

Google’s New Strategy and How It Affects Aggregators

As someone who has had a lot of success using SEO as a tactic to grow companies (for Apartments.com, Grubhub, and Pinterest it was the dominant channel for new users), I get asked a lot of questions about SEO as a strategy today. Andrew Chen’s Law of Shitty Clickthroughs states that all acquisition channels have a shelf life and decay over time. SEO has had by far the longest shelf life of any major internet channel. It has been a stable platform (unlike Facebook), it consistently grew itself, and it was supported by a very strong business model that could drive revenue growth for Google for well over a decade, so Google didn’t need to monetize all of the free traffic they distributed to other companies. Perhaps a more elegant way of explaining this is that since organic search exists to serve user needs, not advertiser needs, it was a more sustainable acquisition channel, precisely because it was not built to be a channel. People never wanted ads, more email, etc. like other acquisition channels. On Google organic search, people do want answers.

It’s this last statement remaining true amidst a platform shift to mobile that will mark the inevitable decline of SEO as a channel for user acquisition. Ben Thompson declared Peak Google a few years ago as a company. Why he was wrong then is why I am right today by declaring we are now past Peak Google as an acquisition channel. To understand this, you have to understand Google’s strategy. Google’s search engine is driven by optimizations that help its users. Ben Thompson does a good job explaining how this plays out with publishers. If you, like Google, have been analyzing its users for the last few years, you may have learned a few things. The first is that the majority of them are on mobile, where their time is more limited, their connections are (still) slower, and there is the threat of an app replacing frequent queries.

What Google is seeing is that their users no longer want to click ten blue links. They don’t have the time or the bandwidth, and there are now a plethora of competitors in the form of apps for many of those queries. The form factor is dictating the optimal user experience, and forcing Google to evolve. Users want an answer, and they want it immediately. So, that is what Google is doing. If you type a question into Google with a clear answer, there’s a good chance Google will just answer the question instead of recommending a site for it. We’ve all seen that. What’s more interesting is what Google is doing when there isn’t an answer, and the solution is to provide options, or what they would likely call a discovery experience. Recipes is a great example. Where you used to be treated to a bunch of “21 best recipes for X” pages, you now just see the recipes as results.

One would think these queries play perfectly into Google’s existing strategy of ten blue links. But Google knows consumers don’t want to click back and forth onto multiple sites to see each site’s recommendations. They want Google to surface up the options directly. And that is what Google is now doing. Take any top query category, and you will see Google replacing links with either answers or options showing up directly in search.

Whereas the top strategy historically for these “options” queries was to build an aggregator and rank at the top by having the most options, Google is now stating that it should be the only aggregator. In the same way Ben Thompson described the squeeze between Google’s demand for fast loading pages and banning of obtrusive ads on Chrome, Google’s search policies are doing the same for aggregators who do well on organic search.

How is Google doing this algorithmically? Google has started to seriously enforce two new policies over the last couple of years in their algorithm: internal search and doorway pages. For internal search, Google says:

Don’t let your internal search result pages be crawled by Google. Users dislike clicking a search engine result only to land on another search result page on your site.
Source

For doorway pages, Google says:
We have a long-standing view that doorway pages that are created solely for search engines can harm the quality of the user’s search experience.
Source

On the surface, these rooms seem sensible. If you continue to read to the end of the doorway pages update, you may start to see a problem:

  • Is the purpose to optimize for search engines and funnel visitors into the actual usable or relevant portion of your site, or are they an integral part of your site’s user experience?
  • Do the pages duplicate useful aggregations of items (locations, products, etc.) that already exist on the site for the purpose of capturing more search traffic?
  • Do these pages exist as an “island?” Are they difficult or impossible to navigate to from other parts of your site? Are links to such pages from other pages within the site or network of sites created just for search engines?

Reading these two together, what Google is saying is that creating pages indexing your search result pages is a bad experience, and creating new pages that replicate your search experience in a different way for search engine visitors is a bad experience. These two rules effectively penalize any presentation of your inventory of content. How do they tell if these pages are created solely for search engines? It’s similar to how they are detecting bad ads: they use Chrome browser data. So, the guideline for an aggregator who would like to show their content to Google is simple: give us your content, and we’ll aggregate it, or play in the tiny space that is a non-internal search page and a non-doorway page. That effectively means creating a page with unique inventory that does not look like search, yet receives traffic to the page from other sources besides Google. There is a window there, but it is a small one.

While updating the algorithms for these rule changes, Google is figuring out how to be the aggregator in many categories now, and over the next ten years will go down the list of every top searched category and figure out exactly how to do that. To do this, Google will either build, buy, or partner with existing players. Let’s take a look at some of these examples.

The Build Case: Google Local
Google tried to acquire Yelp to build local listings and relationships with local businesses. When Yelp refused, Google built out Google Local on top of Google Maps and Google Search, and now has direct relationships with thousands of businesses managing their information directly with Google. This was not very difficult when you have the dominant search product and the dominant maps product to build on top of. When you search a local query, you see no ads, just Google Local above all other search results.

The Buy Case: Google Flights
In July of 2010, Google acquired ITA Software, forever depressing the market caps of many travel-related internet businesses. ITA powered flight search and pricing information for many top online travel agencies. As you might have guessed, that data now appears on Google directly for free. Google is monetizing that space, and looks to moving from a pay per click model to a more transactional model over time.

The Partner Case: Google Events
Late last year, Google launched a dedicated section when people search for events that aggregates events from third party sources, including Eventbrite. Third parties give their inventory to Google, and Google ranks the events on its own. You cannot rely on your business to be in this category. Google will likely partner if:

  • There is no dominant player they can buy
  • Supply is fragmented and data unstructured
  • There are multiple companies willing to implement specific markup to appear in these discovery experiences
  • The area is not one of the leading categories for monetization for Google today

What do you do if you’re affected?
If you are an aggregator, and Google is moving into your space, it changes your SEO strategy entirely. Whereas you used to create and optimize pages that aggregate inventory for popular queries e.g. “san francisco food delivery” for Grubhub, those pages will now be de-valued as Google replaces those listings with its own aggregator. The best solution to this problem is to shift your strategy to distribution of your individual listings, so that you can outrank competition inside Google’s new discovery experiences. These pages usually need to updated to AMP formats, as that it what is powering all of these new discovery experiences inside Google search.

Many companies will attempt to opt out of this strategy, thinking it will help them preserve their current rankings if they delay or hurt Google’s ability to build a compelling, competitive aggregator to their own. This is unlikely to work. If there are any competitors to your aggregator, game theory will incentivize one of them to partner with Google to steal share. If you are a monopoly and opt out, it incentivizes Google to build a competitor that will threaten your monopoly, you will receive a lot less traffic that also threatens your monopoly in the interim, and Google can dedicate a lot of resources to a competitive product over many years. This appears to be working with Google Local vs. Yelp. Certain companies have been able to thrive despite these types of platform changes in the past by building loyal audiences with high switching costs, like Amazon with Amazon Prime when Google launched Google Shopping.


Google’s strategy has changed, though it will take years to propagate throughout every popular category of search queries. You can’t fault them for this change, as it is the right response to cater to their users. Right now is the right time to understand their strategy and to best position your company for its inevitable rollout. Gone are the days where you can rely on Google for a steady stream of free customers without putting in that much effort. You need to think strategically about where the company is going, if there are still opportunities where your content can attract Google visitors, and how you maximize the now declining opportunity.

Thanks to Randy Befumo for helping me through an early draft of this.

Currently listening to Challenge Me Foolish by μ-ziq.

What Are Growth Teams For, and What Do They Work On?

This blog post was adapted from a presentation I did recently. Hence, slides. Don’t say I didn’t warn you.

I receive a lot of questions about growth teams. Naturally, there is a lot of confusion. Is this marketing being re-branded? Who does this team report to? What is the goal of it? What do they actually work on? When do I start a growth team for my business?

The purpose of growth is to scale the usage of a product that has product-market fit. You do this by building a playbook on how to scale the usage of a product. A playbook can also be called a growth model or a loop.

The first question you should ask before asking about growth is if you have product-market fit?

The traditional definition above is qualitative, and if you’re like me, you like to have data to answer questions. The best way to get that data for most businesses is to measure retention.

The best way to identify the key action is to find a metric that means the user must have received value from your product. The best way to understand the frequency on which you should measure that metric is how often people solved this problem before your product existed. Let’s look at some examples from my career.

For Pinterest, a Pinner receives value if we showed them something cool related to their interests. The best way to determine if the Pinner thought something we showed them was cool is that they saved it.

For Grubhub, this was even easier to determine. People only receive value if they order food, and when we surveyed people, they ordered food once or twice a month (except for New York).

Once you have a key metric and a designated frequency, you can graph a retention curve or a cohort curve. If it flattens out, that means some people are finding continual value in the product. But that is not enough.

Brian Balfour has a great post on this, which he calls product-channel fit.

If you’ve been around startups for a while, you might remember this tweet from Paul Graham. It talked about the fastest growing startup Y Combinator has ever funded. It is a graph of revenue growth from $0 to $350,000 per month in just 12 months.

The startup was Homejoy, an on demand cleaning service. Investors liked this graph, so they gave the company $38 million to expand.

20 months later Homejoy shut down. From a post-mortem of the company, I highlight the following quote.

If you discounted to get to product-market fit, you didn’t get to product-market fit. Product-market fit is not revenue growth, it’s not growth in users, it’s not being #1 in the App Store. Product-market fit is retention that allows for sustained growth.

So, I though product teams were in charge of creating a product people loved, and marketing teams were in charge of getting people to try the product. What changed?

What changed is an acknowledgement of what actually drives startup growth. There are three main levers. Phase I is simultaneously the most important and the least understood. In Phase I, you change the product to increase its growth rate. Some changes include improving onboarding, helping the product acquire more customers through activities like virality or SEO, incresing the conversion rate, et al.

These initiatives are “free” in that they don’t require an advertising budget. Their cost is the opportunity cost of a product team’s time. They are measurable in that you can create an experiment and understand the exact impact of the change. They are also scalable in that if you make a change that, say, improves your conversion rate, and it has a certain amount of impact, it likely will have that same impact tomorrow, weeks from now, and potentially even years from now.

The other two phases are what we traditionally think of as marketing. Performance marketing initiatives, like buying ads on Facebook or Google, are also measurable and scalable, but scale with an advertising budget. Brand marketing usually requires an even larger ad budget, and is harder to measure or scale. The time frame over which brand marketing works takes years, and can be hard to confirm. If you do create a PR campaign or a TV ad that seems to work on a more immediate time frame, it can be hard to scale. that is because brand marketing always requires new stories to keep people’s attention.

This is why marketing can’t be in charge of all growth initiatives. They don’t have access or capability to the most important ones. They might know they need to improve the site’s conversion rate or get more traffic from referrals, but they don’t have access to the product roadmap to get them prioritized appropriately, and if they do get engineering and design help, they don’t have the expertise in working with them to build the best solutions.


Perhaps what’s more important to understand in the difference between marketing and growth is how the traditional marketing funnel changes with startups. Above is the traditional marketing. This model is based on the old school model of product development before the internet in spending a lot of money to make people want things.

Startups by definition should be making things people already want. When you do that, you can invert the funnel and focus on people that already want the product or people that are already using it. This is more effective on a small (or no) budget.

When you translate that into tactics, you see how product-driven growth initiatives dominate the top of the list of priorities. It does not mean you won’t work on performance marketing or brand marketing, but that they usually become important later on in a product lifecycle as an accelerant to an already sustainably growing company.

So I spent a lot of time explaining why growth is different from marketing. How is different from product?

Growth teams don’t create value. They make sure people experience the value that’s already been created.

The most common examples to start a growth team to address are:

  • improving the logged out experience (for conversion or SEO)
  • sending better emails and/or notifications
  • increasing referrals or virality
  • improving onboarding

SEO and onboarding are harder places to start because their iteration cycles are much longer than the other areas.

Growth teams don’t start by finding mythical VP’s of Growth to come in and solve all of your problems. They are usually started by existing employees at a company (or founders) that really understand the company and what’s preventing it from growing faster. They report to their dedicated functions, but sit together to focus on problem solving.

To find out which area you work on after you have the team, you have to analyze the data. For example, at Pinterest, they originally wanted me and my team to work on SEO. What we saw was that while there was a lot of opportunity to get more traffic via SEO, a bigger issue was the conversion rate from that traffic. So we decided to work on conversion instead.

Then we had to figure out what to work on. Jean, an engineer on the team, had recently run an experiment that gave us a key insight. So, we said, we could use this same modal when people clicked on Pins. Clicking on a Pin could show enough interest in Pinterest for you to want to sign up.

The other thing people did when they liked what they saw scroll to see more. So, we decided to try stopping them where we stopped the Google crawler, and asking them to sign up then.

It took Jean two days of work to launch this experiment, and it resulted in a much bigger impact than expected.

So that’s an example of finding a conversion issue in you data, and putting together a really scrappy experiment to try to improve it. What else can growth teams work on? Here are some examples from my time at Pinterest, and some best practices we’ve learned.

Usually, the biggest area a growth team focuses on improving is retention. That’s right; growth teams are not just about acquisition. Retention comes from a maniacal focus on improving the core product, which I define as core product, not growth, work. Where growth comes in is reducing friction to experience that core product. Simplifying how the current product works usually has much more impact than adding new features. New features complicate the product, making it harder for new people to understand.

So how do you simplify the core product? Well, you have to have data to understand what people do, and pair it with qualitative research to understand why they do it. We spent countless hours at Pinterest putting laptops in front of non-users watching them sign up for the product to figure out why people didn’t activate.

At Grubhub, data pointed out that Grubhub was an S curve when it came to both conversion and retention. This graph is a (now very old) graph of conversion rate in Boston based on how many restaurants Grubhub returned when you searched your address. After 55 results, conversion rate essentially doubled for new and returning users.

Qualitative research gave us different insights. When we asked users why they didn’t use Grubhub more often, they would say, “it’s expensive.” We thought that was weird, because Grubhub wasn’t charging them anything. What they meant is that delivery was expensive due to minimums and delivery fees. So, we went back to our restaurants, convinced a few to try lowering their minimums and fees to see if increased volume could make up for lower margin. When it did, we creates case studies to help convince other restaurants.

At Pinterest, we simplified the signup and onboarding flow. What used to be a flow that required five steps was now three with one of them pre-filled and the other two optional. What we did do was introduce friction that we knew made it more likely a Pinner would find content they care about. This was asking them which topics interested them before showing them a feed of content.

We also realized that the more content people see, the more likely they will find something they like, which will lead to retention. So, we removed content around Pins that was non-critical, like who Pinned it to what board and how they described it. All of these increased activation rates.

We also contextually educated people on what to do next when they were onboarding. There is a common saying that if you need to add education to your design, it’s a bad design. It’s pithy and sounds smart, but it’s actually dangerous. My response is that a design with education is better than a design that doesn’t educate.

Search engine optimization has been a really great lever for organic growth for every company I’ve worked on in my career. It’s not for every business though. People need to already be searching for what you do. That alone isn’t enough though. You need to be an authority on the subject, which Google determines by relevant external links to your domain and your content. You also need to be relevant to what was just searched.

We worked on improving both of these at Grubhub. When Grubhub launched new market, by definition we weren’t locally relevant yet. So we would go to local blogs and press outlets and tell them we were launching there, and that we wanted to give their readers $10 off their first order. All they had to do was link to a page where the discount would auto-apply. After a while, that page would have enough local links so that even though the promotional discount was over, it would still rank #1 for the local delivery terms e.g. “san francisco food delivery”.

For relevance, Grubhub knew which restaurants delivered where, their menu data, and reviews from real people. So we aggregated them into landing pages for every locale + cuisine combination e.g. ‘nob hill chinese delivery”.

We applied the same landing page strategy at Pinterest. While Pinners had created boards on their favorite topics, it was one person’s opinion on what was relevant for a topic. Pinterest has repin data globally for every topic, so we knew what the best Pins were across the Pinterest community. So we created topic pages with the best Pins, and they performed better than individual boards on search engines and with search engine users.

We also worked a lot on emails and notifications on the Pinterest growth team. Emails are a key driver of retention. They won’t solve your retention problem, but they will certainly help if you do them right. At every company I have been at, people hated email and didn’t want to send them to their customers. When they finally did, they saw lifts. You are not your customer. You get more email than they do. Emails help them if they’re connected to the core value of why they use your product. Emails are not helpful if they’re pushing a marketing message.

At Pinterest, I made this mistake. I set up campaigns with emails that explained all of the things Pinterest could do. People don’t care about what Pinterest can do. They care about seeing cool content related to their interests. We needed to stop sending email like a marketer, and start sending email like a personal assistant. So we replaced those emails with popular content in topics of interest for each Pinner, and our retention increased.

Then, we built a system around it. Each Pinner likes different content, at different times, and different amount of it. So we learned for each Pinner what content they liked, when they liked to receive emails and notifications (based on when they opened them), and how much they liked to receive based on testing different volumes and seeing open rate impacts.

If you’re testing emails and notifications, you can test manually first, then automate and personalize. What I have learned at Pinterest and Grubhub is what seems to be worth testing. At Pinterest, one engineer tested 4,500 different subject lines, resulting in hundreds of thousands of additional weekly active users. Around the same time, we spent three months redesigning all of our emails, and it had no impact on usage.

A common issue I see with growth and marketing teams is they think that emails and notifications can only have positive impact. This is not true. You have to measure the lift in usage vs. the unsubscribes (and the impact of an unsubscribe) and app deletions. Those will impact usage, and you need to know how.

Growth teams have a clear purpose, and that purpose makes sense only if you have first found product-market fit. Once you have that, you will find traditional product and marketing lacking in their ability to help scale usage of your product. That’s where growth teams come in. Growth teams use data and qualitative research to help understand the frictions that prevent more people from finding the value in your product. That can mean acquisition, but it can also mean reducing friction in the core product, working on conversion or onboarding, or finding ways to remind existing users about the value you’re creating. If you have questions about growth teams, don’t hesitate to reach out to cwinters@greylock.com.

This presentation was made in conjunction with @omarseyal, who is awesome.

Currently listening to Everybody Works by Jay Som.

Four Strategies to Win Big with Low Frequency Marketplaces


Frequency creates habit which creates loyalty which creates profit. Uber and Lyft are successful because consumers need to get from A to B multiple times a day, forming habits that lead to long (and high!) lifetime values. Grubhub similarly benefited from people eating more than once per day.

But there aren’t that many business opportunities that have daily — or even weekly — frequencies. And those spaces have become very competitive. For example, how many food delivery companies can you name? Now add in groceries or meal kit cooking companies. All that just for the “eating” use case.

What if the natural frequency of use for a transactional business is low, like buying a house, selling a car, or booking a trip? How do you create a successful business if ideal frequency is quarterly or yearly or even once every few years? You would be unlikely to create a habit or loyalty, much less get the customer to remember your brand name. That is usually the case. If you don’t create loyalty, then you usually have re-acquire consumers when the need eventually arises again. This hurts customer acquisition costs and lifetime value. This fact makes building a successful business with low frequency extremely difficult.

With a low frequency business, you usually need to have a high average selling price to make up for the lack of frequency. While an order on Grubhub may cost you only $25, the average transaction size on Airbnb is hundreds of dollars. But a high average selling price alone is not enough to become a massively successful business. I’ve seen four distinct strategies for how to thrive in low frequency marketplaces. They all revolve around being top of mind when the transactional need occurs, no matter how infrequent that need is. I’ll start by talking about the most common approach, and then lead into some that are actually more valuable and defensible.

The Expedia Model (AKA SEO)
Companies that pursue this model: Thumbtack, Expedia, Apartments.com, WebMD

My first job was at Apartments.com. We were a classic low frequency marketplace. People search for apartments at most once a year, and there isn’t a whole lot of value you can provide in between apartment searches. So what did we do at Apartments.com? If you do not create habits or loyalty with initial use, users go back to the original way they solved the problem last time. Where do people go where they are searching for help renting an apartment? Usually, Google. So, the Apartments.com strategy was to rank well organically on Google so when people did search again for an apartment, they’d be likely to see us and use us for their search again.

SEO can be a very successful strategy, but the entire company has to be geared around success on Google. This strategy is also susceptible to platform shifts, like Google algorithm changes or Google deciding to compete with you. It also tends to shift companies toward portfolio models at scale. This is why Expedia owns Hotels.com, Orbitz, Hotwire, Travelocity, and Trivago, and why Priceline owns Booking.com and Kayak. When you rank #1 for your main keywords, the only way to grow is to own the #2 and #3 spots as well.

The Airbnb Model (AKA Better, Cheaper)
Companies that pursue this model: Airbnb, Rent The Runway, Poshmark

Sarah Tavel wrote a post about products that are 10x better and cheaper than their alternatives. You can definitely pursue this strategy even if you have low frequency. Airbnb was significantly cheaper than hotels, and many people, once they experienced Airbnb, found it a better experience as well. It was a more unique listing, in a “more real” part of the city, and they had a connection to a local. So, even though people only travel once or twice a year on average, when they do, they remember the Airbnb experience and start there directly instead of on Google, competing with the SEO behemoths of Expedia and Priceline.

Finding this level of differentiation in different industries is not easy, but worth contemplating. Airbnb is not the only startup that has entered a crowded space and grown rapidly by figuring out how to be 10x better and cheaper. RentTheRunway allows you to access high quality fashion without the high price, and without storing it, because dressing up is increasingly a low frequency occurrence.

The HotelTonight Model (AKA Insurance)
Companies that pursue this model: HotelTonight, One Medical, Lifelock, 1Password

There are certain businesses that are needed infrequently, but when they are needed, they are needed with great urgency. Example spaces include urgent care, being stuck in a random city unexpectedly, and fraud alerts. The key here is that someone keeps the app or account live despite a lack of usage because the fear of when it might be needed is so great. This is a hard strategy to pursue, but once the value prop is established, these companies remain sticky despite their lack of frequency.

The Houzz Model (AKA Engagement)
Companies that pursue this model: Houzz, Zillow, CreditKarma

Contrary to what many might think, keeping users engaged in a low frequency business is indeed possible: the key is a non-transactional experience. Many of these approaches have a “set and forget” component to them where they reach out with pertinent information in a more frequent way. Zillow is the first example I can remember that utilized this strategy. Even when not actively looking for houses to buy, Zillow kept users engaged by valuing their existing homes via the zestimates. CreditKarma reaches out with alerts and monthly credit check updates.

Houzz is a great example that is more recent. People remodel and redecorate their homes infrequently, but they are inspired more regularly. Houzz has a great product that shows home inspiration that can be saved and discussed, and when needed, but much more rarely, transacted.This is a product people engage directly with in instead of having to have content pushed to them

For this strategy to work, you essentially build a second product that enables frequent engagement — not a transactional product. Engagement strategies for low frequency marketplaces take advantage of an inherent human desire to stay up-to-date on things important to them. This won’t work for all industries. We actually tried this at Apartments.com, but were not successful because renters don’t care as much about investing in their living situation as homeowners.

A common confusion is that loyalty programs are an example of this. What loyalty programs usually do is increase frequency or target users that have high category frequency, like business travelers in the travel segment, rather than create loyalty from infrequent users. It is still a very valuable strategy, and I have blogged about loyalty programs if you want to learn more.

Of the four models I wrote about above, you will notice that not one of these is a brand model. Many of the sites listed in the SEO model have spent hundreds of millions of dollars building brands. Yet most travel searchers still start with Google. Brand is an extension of the Airbnb model, not its own strategy. If the product doesn’t deliver on a differentiated experience, brand building usually does not create loyalty.

So, if you’re building a low frequency business, do not dismay. There are many paths to still becoming a very large and differentiated business. These strategies are difficult but very rewarding if they are executed well.

Currently listening to Take Me Apart by Kelela.

The Real Value of PR for Startups

Many startups get obsessed with press. Not just CEOs, but also employees, friends of employees, and investors. Press, like paid acquisition, can be a drug though. I’ll talk about some of the ways the startup ecosystem perceives press is bad, what it is actually good for, and some outliers.

Press Abuse
First, let’s talk about some of the abuse of press by startups in the past. Business insider has a great recap of some of the issues Evernote faced in the couple years before their CEO stepped down. I’ll highlight this quote in particular.

“There was a feeling that we were working on the wrong priorities,” a former employee said. “It was clear the motive was to just continually drum up press. They had no idea how to optimize and improve growth.”

From The inside story of how $1 billion Evernote went from Silicon Valley darling to deep trouble

Startups frequently correlate press with growth, and think that they need to create new stories to drive growth. Then they start changing their entire product roadmap to drive press instead of value to customers. Startups should not look for PR to drive growth. PR doesn’t usually drive growth. And when it does, it’s a bump, not an engine. In startups, you want to build growth engines.

This is arguably a better problem than the second way PR drives bad acting in the ecosystem: founders getting addicted to the attention of press. It feels great when there’s a picture of you in the Wall Street Journal and your mom sees it and your dad’s friends. And founders frequently want to create that feeling. It’s as if appearing successful is more important than being successful. This can be a dangerous pattern.

Another problem with press is that it helps competitors understand exactly what you’re doing. I’ve talked in the past of the advantage of being a silent killer and why seeking press attracts unneeded competitor attention. I won’t repeat that here.

What Press Is Good For
A common refrain I tell startups is that press is good for three things: investor interest, employee interest, and links for SEO. I’ll break those down a bit, and talk about why they matter. First, let’s talk about investors. Very few businesses make sense for venture capital, but when they do, they increasingly need larger amounts of it. The venture business has extreme cases of FOMO. One deal can make all the difference. So investors rely on a lot of signals, and the tech press is one important one they look at not only to find new business to research, but to also get comfort in their interest in certain businesses. When Grubhub was raising its Series C from Benchmark, we got written up about our iPhone app on Techcrunch. For Grubhub’s growth, it didn’t matter at all. But it looked great for our fundraising.

Likewise with investors, potential employees rely on press to hear about potential new places of employment and to feel comfortable they are making a good move. Recruiting outreach will have a lower conversion rate if the person has never heard of your company, or if the person’s network hasn’t heard of you when they ask around. This can be overcome, especially if you’re well connected or can tout amazing metrics privately (like WhatsApp), but my guess is WhatsApp always had a harder time recruiting than Snapchat.

Lastly, press is very helpful for SEO. SEO is about relevance and authority, and links from press are a great way to increase authority in the eyes of search engines. Not only can press drive a quantity of links, but the quality of links are probably the highest most startups will receive. So many people link to news publications that they have among the highest domain authority on the internet. Now, getting one of these links won’t make you rank #1, but it’s part of a healthy SEO strategy.

When Can Press Be an Engine of Growth?
Now that I’ve set an argument that press doesn’t usually drive growth directly, I want to talk about some outliers of when it does. Press can drive growth if a certain story about your startup is self-perpetuating. Then, the volume of articles written about you happens in such a regular cadence that readers eventually act on it. I have seen this happen in two major startups over the last ten years in Twitter and Uber. With Twitter, it received such a tremendous uptake from journalists that they kept writing stories about it. This drove their readers to try it, and it became a sustained vehicle for organic acquisition (unfortunately with a low activation rate, but that isn’t press’s fault, but the product). Uber’s growth via press was a little more sculpted. They were able to create a story about the business of Uber fighting against corrupt incumbents and governments that were preventing a better product from succeeding. It was almost a movement. For years this created a headline about Uber almost every week as they faced this resistance in almost every city in which they launched.

I want to highlight that these are anomalies, and it’s unlikely you will be able to create sustained growth directly from press in your company. But that doesn’t mean press is worthless. Just remember what it’s good for, and make it work for you for those interests.

Why Everyone Link Builds, or Why You Sometimes Do Things When You Don’t Know If They Work

I was talking to an analyst about how SEO works, and we inevitably got to the authority side of SEO. I started talking to him about how many companies spend a lot of effort trying to get external links to their site to build authority. It’s not something we have to worry about at Pinterest as our authority is super high naturally, but most companies do not have that luxury. The analyst, being a good analyst, asked how you track effectiveness of link building as a program. My answer surprised him: most of the time, you don’t.

Sure, sometimes you can see a correlation between link building and average weighted rank improvement, and maybe you didn’t make any improvements on the relevance side during that time. But, while you can experiment with SEO in relevance changes, it is pretty much impossible to experiment with link building as it works at both a domain and page level, its impact is felt over such a long period of time, and there are almost always so many other factors one can’t control, namely from competitors for those same search results.

So, he then asked, how do we know it works? The answer is: we usually don’t. So, the next question of course is, if it’s a pain to do and we don’t know it works, why do we do it? This question can be generalized to almost any competitive question via game theory. To really hammer the point home though, I actually used the climate change example.

In climate change, their are four scenarios and two dimensions. The first dimension is whether or not it is actually happening (or whether it’s man made), and the second dimension is whether we do something about it.

Next, you examine the outcome of each scenario to determine the outcome if that’s the box you pick.

As you can see here, in every box you are fine, except for the upper left. Sure, you might have wasted quite a bit of money and slowed the growth of some businesses, but none of that compares to possible catastrophe. Even if you think it’s a waste of time, the risk is so great if you’re wrong, and the answer reveals itself over such a long period of time, the obvious answer becomes to assume it is true and invest in fixing it.

Now, let’s apply this to a much less risky scenario of link building.

While we aren’t saving the world when we work on SEO, from a business perspective, the risk is just as great. In one belief, you lose, and in all other scenarios you don’t. If your competitors are smart, they all do this exact analysis and come to the same conclusion: to invest in link building. This cannot be a prisoner’s dilemma either, as one company always outranks another, and links occur organically that presumably change the rankings.

A Primer on Startup SEO

As I advise startups on growth, one of the most common questions I receive is “should we be working on SEO?”. At this point, I tend to remind them of the three phases of scaling startup marketing or show them Andrew Chen’s great post on the few ways to scale user growth. In today’s mobile-first landscape, with the limited scale of the App Store/Google Play, the answer is actually a bit more nuanced. So I created a guideline on how to answer this question for any startup.

Step 1: Keyword Research
The first question to answer when thinking about SEO for your business is “what should my business show up for?”, which really is asking the question, “what is my business about?”. Unlike branding, where the point is to synthesize that answer into as few clear words as possible, keyword research is about generating as many answers to that question as possible. These will be your potential keywords. I like to use the framework: who, what, when, where, how to answer this question. Let’s take the example of GrubHub:

Who: GrubHub, restaurant names we represent
What: food, delivery, menus, pizza, thai, indian, chinese
When: breakfast, lunch, dinner, late night
Where: every city, neighborhood, zip code, college covered
How: online ordering, mobile app, iphone app, android app

Take these words, combine them all into new keyword combinations in Excel e.g. “late night pizza delivery berkeley”, and check to see how much search volume they have. To do that, use the Google Adwords Keyword Planner. Click “Get search volume for a list of keywords”. Take the keywords from the exercise above and paste them in. Click “Get search volume”. Click on Keyword Ideas tab. You can also do this process again to find new keywords by clicking “Search for new keyword and ad group ideas” at the beginning.

You can now see how much search volume is closely related to your business. Search volume determines how much your prioritize SEO for your business. Now, there is no hard and fast rule for how much search volume there needs to be for you to get excited about it as an opportunity. It depends on purchase size, how much you want to grow, etc. Generally, I recommend taking all of those keyword’s search volume, assuming you can get a very small percentage of it to your site e.g. 1%, and seeing if that would make an impact on your business.

Now that you have an idea of the search volume for your business, you need to know how competitive that real estate is. Fortunately, Google estimates how competitive they think each keyword is next to the search volume estimate. It lacks the granularity I’d like, but it’s a good starting point. What I do in addition to looking at this gauge is do some spot searches and see what the results look like. Here, I’m looking for two things:
1) Are the results primarily businesses you would consider competitors or blogs?
2) Are the pages that are showing up well optimized for the keywords you searched or not? Are they dedicated landing pages or home pages?
(I’ll go into more detail on how to measure this a little later in this post)

Step 2: Picking an SEO Strategy
Now, you should have an answer to two questions:
1) Do my keywords have a lot of search volume?
2) Are those keywords competitive or not?

This creates four scenarios:

High Search Volume/Low Competition: Make Priority
This is a rare opportunity, and you can build a great business just off of SEO here. You should make SEO a priority and one of your primary growth strategies.

High search volume/High Competition: Play for Long Term/Make Core Competency
This means the ROI is there with SEO in the long run, but it will be hard to get it as a startup. You will need to organize the entire company around SEO to win, by making SEO a core competency of the company.

Low Search Volume/Low Competition: Content Marketing
This is common with startups that are creating new products or product categories, especially in mobile apps. They do not have enough demand yet. The general approach here is to expand your keyword target to the broader industry that does have high search volume, and pursue a content marketing strategy around those keywords that mentions your product/service occasionally in them.

Low Search Volume/High Competition: Ignore
Feel free to ignore SEO as a strategy here unless something changes.

This allows us to build a 2×2 to express SEO strategy options:

Step 3: Actually Working on SEO

Assuming you’re not in the low search volume/high competition bucket, you’ll want to start figuring out how to work on SEO. To do that, it’s helpful to make sure we define what we’re doing. Search engine optimization is the process sites use to appear in the organic results of search engines. To have a process, you need to understand what search engines do.

1) Search engines use crawlers to discover pages across the web.
2) They read any content they can find (mostly text)

So, in order to succeed, you need to be discoverable and readable. Once your page is discovered, search engines determines the authority of the page
Once your page is read, search engines determines what the relevance is for certain searches.

Determining Relevance: On-Page Factors
So how do search engines determine relevance? Well, here’s a rough hierarchy. They look at the title tag of the page first then the H1’s and H2’s, thing that typically indicate importance in HTML. Then they read normal text, and they look at what the URL says. They also look at this page compared to all the pages in their index and see how unique this page is compared to the rest. Search engines prefer unique content. They also look at the last time the page was updated. Frequently updated pages are seen as more reliable to Pinterest. They also look at the # of links on the page. A page with a ton of links is associated with a worse user experience and having less relevance. They also look at where the keywords are on a page. Google breaks up the page into header, footer, sidebars, and content area. Keywords in the content area are weighted higher. They also look at the # of content types. A page with text, video, and images is seen as better than just a page with one of those. They also look at the # of ad blocks on the page. A page with a bunch of ads is seen as less relevant. I know that sounds like a lot to pay attention to, so just keep this short list handy:

  • Title Tags
  • HTML Tags
  • Text
  • URL of the page
  • Uniqueness
  • Freshness
  • Number of Links on the Page
  • Keyword location on the Page
  • Diversity of content types
  • Number of ad blocks on the page
  • Make no mistake about it. This is mostly engineering work. You have to be messing with your site to get it to rank better. Mostly, this means creating pages specifically for keywords you want to target, and optimizing the above for these pages. I have seen so many startups think they can cover SEO by hiring a marketer to manage it. Unless they get engineering help, it will not work.

    Note: this is what you check to answer if your keywords are competitive in Step 2.

    Determining Authority: Off-Page Factors
    So, how do search engines determine authority? Well, the main two things are quantity and quality of external links to the page and domain. Search engines see links as votes, so if another site links to you, that’s a vote that you’re an authority. Now, not all votes are ranked equal. A link from the San Francisco Chronicle will be worth more than a link from my blog. They also look at how other sites link to you. So, the anchor text is very important. If a link says home decor, that will help more than a link that says click here. They also do look at internal links within a site. So, us linking to something from our home page indicates that we think it’s very important, where as a link from our Help page is not as important. Search engines also look at the data they accumulate about a page. So, when a page gets clicked from Google, they look at the bounce rate. When Google shows a page in a search result, they also look at its click through rate. They also look at which parts of the page people link from. A link from the content area of another page is worth more than a footer link. They also look at the diversity of link types. A page that gets links from blogs, news sites, and social media will be better than just a bunch of links from blogs. Too much detail again? Don’t worry; I have you covered with another list:

  • Quantity of external links pointing to page/domain
  • Quality of external links pointing to page/domain
  • Anchor text of links
  • Internal links pointing to the page
  • Metrics from search engines (bounce rate and click through rate)
  • Area on page of internal/external links
  • Diversity of link types
  • If you’re building a good company, this is mostly public relations and business development work. Also, if you’re working on content marketing, the quality of the content alone can drive links, which is why you see every company trying to push their infographics everywhere. Widgets have traditionally been a strong strategy here that may be waning in importance.

    Appendix: Tools at Your Disposal
    Now, search engines give you some tools to help you do this. So, I’ll describe them.

    Google/Bing Webmaster Tools: These destinations give you a host of information on keywords you rank for, crawl rate, errors etc.
    Meta Tags: By default, search engines will use your title tags and meta descriptions to populate how your listings appear on their sites
    Sitemaps: This is a way to send search engines every page you want them to index instead of waiting for them to find a link to it. No guarantee they’ll index all those pages, but they’ll look at them.
    Nofollow Tags: With the explosion of user-generated content and social media, spammers started flooding these sites with links to rank on search engines. Given that it’s very hard to monitor all that content, search engines allow sites to add rel=nofollow to outbound links, saying you can’t vouch for the site you’re linking to. Pinterest uses nofollow to external links as do most other social media sites.
    Canonical Tags: another cool tag. As we said before, Google likes unique content. But Google may figure out how to access the same content from multiple URL’s. If that happens, when Google finds a duplicate version of a page, you can add the rel=canonical tag in the head to indicate that if this page gets a link, use it for this other URL with the same content.
    Hreflang Tags: helps tell Google which version of a page to show users in different languages and countries.
    301 redirects: URL’s change all the time. But if a URL changes, normally that would be considered a new URL that needs to generate its own authority. If you 301 redirect the old URL to the new URL, Google will transfer some of the authority of the old URL to the new URL.
    robots.txt: Search engines obey operatives in your robots.txt file or in meta robots on which pages to crawl or index.
    Rich snippets: This will show different content under your listing, like star ratings and other meta data to help your listing stand out.