How We Initially Did SEO at Our Startup SeatGeek (Distilled Contest)
When I saw Distilled was doing a contest for Startups I got a cocky smile on my face. Or as T.I. said “I’m the opposite of moderate, immaculately polished with the spirit of a hustler and the swagger of a college kid” – in this case the swagger of a future SearchLove attendee who will be shouting “I won the searchlove competition”.
I had been eyeing SearchLove but did not have the budget to attend, and knew this was my ticket to what some are calling the greatest SEO Conference ever (no one is saying that, but they could be…I will if I win because I haven’t been to any other SEO conferences).
INTERLUDE for one of our more Creative Links
I probably wrote a little too much for this post, so to keep attention here is one of the more creative link bait pieces we did. When the Earthquake “hit NYC”, a developer on our team (without any request of mine) noticed an impact on page response times because our Amazon AWS servers were all in Virginia, right near the epicenter. We quickly put out a blog post What An Earthquake Does To Page Response Times, which trended at the top of HackerNews and received non-solicited links from PC World, ZDNET, NetworkWorld and the CNN Tumblr all courtesy of @acslater00 and @michaeldauria.
Ok now back to it…
Where We Started with SEO
The image to the right is a screenshot from my original role sheet. You can see we weren’t that sophisticated earlier on
Yep, that’s how my SEO career begin. Well not entirely true, I was the Product Manager for a Local SEO/Profile offering at Vistaprint. But not only are the algorithms different, but I quickly learned that it was one thing to know that getting links was important or on-page optimization was important, and an altogether different thing executing these concepts across a site with tens of thousands of important pages.
Fortunately, although we weren’t explicitly attacking SEO per se at the beginning, we were producing a lot of content, because producing content seemed like the right thing to be doing…and it was our gateway to press coverage. We were going to be the OkCupid Blog of events. Except, everyone likes other people liking them (for the most part) but not everyone likes sports/concerts/theater or cares about price trends within those industries. So we were getting some decent press hits, but the inbound marketing traffic wasn’t there. Our solution, obviously was SE…no it was PPC. Paid search was our big jam, the way I was going to shine at this new company and bring in grips of traffic, earning us stacks on stacks on stacks.
That’s not exactly how things went down. Yes, traffic was growing, but it wasn’t really from search as a channel but more from great partnerships.
What we did right at the beginning:
- Active blog
- Using our data as assets for content and PR outreach
- Flat architecture (for UI reasons, not SEO)
What we did wrong at the beginning:
A lot…even the stuff that was justifiable with data, etc just wasn’t optimal looking back
- Sub-optimal internal linking from our blog (most notably using id parameters on internal links to track things)
- no rel=canonical tags, duplicate content on parameter/affiliate links
- Did not surface top pages as links on the homepage in the form of navigation
- Grossly underestimated what we could achieve from an SEO standpoint (i.e. we targeted the term “sold-out tickets” pretty aggressively which just isn’t a high value term in retrospect
- Focused too much on quick/automated “wins” that would come back to bite us in a post-panda world (i.e. creating a landing page for every event for every city…including Jupiter, Florida…[EDIT: what we did here was have landing pages for all event types for all cities i.e. ‘Jupiter Florida Broadway Tickets”, there was and never will be a Broadway event in Jupiter, Florida]
- On-page factors that showed a noobish understanding of how people search for events (beyond just x team/artist tickets (more on this later)
- No XML sitemaps
- A huge html sitemap
- Link structures for events that started with /event/show/event-id/blah-blah
Where We Went with SEO
After our first Summer Internship program in 2010, and in particular our story on Where A-Rod was going to hit his 600th homerun, we started to truly appreciate the potential of organic traffic with the vast amounts of long-tail terms people were hitting this story on – which wasn’t even intended to be a piece that would bring in search traffic, but rather links and press (which we also got on NY Times and several other major publications. The press was cool, but that was what we intended, it was the organic traffic that caught us off guard as thousands and thousands of people found the page for queries as short as “a rod 600”, as on point as “where does a rod hit the ball” and as long as “if i catch alex rodriguez 600 homerun ball as a fan what happen if i want to sell it”.
Look, none of us at SeatGeek are stupid, I would like to think we are quite smart (Jose excluded…), but sometimes as a young startup – especially one still in seed phase – you need to directly experience the results to justify redirecting time and resources. Not to mention this was the first time around for any of us really scaling out marketing on a product and we were competing with much more “experienced” companies. This one event was the catalyst for our SEO game and it just snowballed from there.
INTERLUDE to show we can do INFOGRAPHICS
We definitely aren’t the most prolific infographiteers on the block, but we did get some great links from such sites as Guyism and Billboard. Plus, we did it with a summer intern on design the one and only Laura Groetzinger, the younger sister of Jack…friends and family kids. Here are the rest of SeatGeek’s Infographics. [EDIT: here is a better one we did for the Final Four]
Ok back to the flow…
SEO in the Collective Company Consciousness
No longer are the days that hundred of pages get rolled out with identical title tags or robots.txt files are edited with no input from me. We now all know that SEO is f*cking important. If you do something stupid, it’s not just Chad Burgess, marketing monkey that is going to be concerned.
It started at the top with Jack and Russ, and with me as I got more and more educated on SEO and had the confidence to really start setting aggressive goals. That was around the beginning of this calendar year, with much props to Mason Pastore, now an MBA candidate at Babson, for being intellectually curious and pushing me on my knowledge. It wasn’t until May though that the dev. team started to be fully on board, and one major turning point was this informal presentation I gave. The slides don’t necessarily speak to it, but it was more just me spitting SEO shit for 45 minutes that got them to have confidence in me.
In fact just this week (no joke) one of the devs jokes to another that “you did really good on-page SEO for ‘delete all but last 100 git tags‘…”
I have done a decent job, though still improving, of cultivating a group of people that see link opportunities in strange places (the Earthquake example above was a great example). We go after links in a variety of fashions:
- Data-driven PR
- Traditional linkbuilding
- Developers Opensourcing
- Developers Writing Content
- Other Team Members Writing Content
- Interns leveraging .edu networks
Ok ok, so I have danced around a lot but did any of our shit actually work? Yes, we finally got those stacks on stacks on stacks…to a degree.
September 2010 Compared to September 2011. This is the last full month of data, but also important dates in SeatGeek’s history as it represents our 1 and 2 year anniversary months. Here are some year over year SEO growth rates.
- We experienced a 645% increase in organic traffic
- With that traffic we were able to achieve a 964% increase in organic conversions
We have made big moves onto page one on informational queries, industry queries and most importantly transactional queries – I have shown a small sample below. Sorry on the quality, just click on them for a better look
The above graph is from Raven Tools, but I had rank tracking on this earlier and at the start of the 2011 calendar year we were 47th in Google.
At the start of 2011 were were actually a little higher for “NFL tickets” at 45 in Google.
In our high-level metrics you can see that our conversion growth has outpaced our traffic growth. The reason behind this was that early on (and not even that early, it took a long time for us to rank for key transaction terms) we were not able to rank page 1 for “tickets” queries like Mets tickets, San Francisco Giants tickets, Foster the People tickets etc. We had to be more creative and find other less transactional keyword modifiers like lineups, dates, years, etc. to get traffic. It wasn’t until just this spring that our first core product pages started to pop into page 1 and we have made lots of progress since then. There are roughly 120 major sports teams resulting in around 240 really key short-tail transactional terms. Of those we rank page one for 20%. Big step up from nothing, but there are still 80% that we don’t rank for. It took us 2 years from the launch of SeatGeek to get where we are and still only 20% there…but no one at SeatGeek would say that we aren’t happy with the progress we have made. Now that we are more savvy, not sandboxed, etc. the moves should and will come quicker. Watch out now…
[EDIT: as of 6/10/2012 we now have over 80% page 1 coverage for major sports tickets queries]
This is the core of it all, the fact that the development team has gotten onboard (along with the founders to allocate dev. time) with this whole SEO thing. They have given us the tools we need to move more quickly than our competitors and optimize with limited programming knowledge and no fear of breaking with bad commits. Jose explains this further:
“SEO is pretty often overlooked in the development community, and it’s hard to quantify it’s benefits to clueless developers. It’s also annoying for a developer to have to update copy manually, and potentially disruptive to the entire workflow. Over the past year, I’ve built an administrative panel at SeatGeek such that non-developers can instantly change page titles, meta tags, and other SEO information on a given page without developer intervention, allowing for a more fluid, iterative approach to our SEO tactics. I’ve also released a plugin for the CakePHP developer community on github – https://github.com/josegonzalez/sham – based on this work to inform developers as to just how important SEO actually is.”
Seriously, I am spoiled with the tools we have to work. It gives us the edge we need against much bigger companies that have full SEO teams and not just a portion of one person’s time.
Hard to get links
First off, our Business Development team has done an impressive job landing partnerships with sites like Yahoo! Sports, Daily News and The Wall Street Journal with integrated ticketing links . There is always the whole debate between partnerships links vs. editorial links, but in a search engine’s quest to find the best sites to surface, why wouldn’t you leverage the due diligence of a site like Yahoo! as a signal of trustworthiness and great user experience? If you comment on one thing, make it this, always a fascinating topic for me.
Of course we have plenty of editorial links:
That is just a few, here are more mentions earned just this summer by our 2011 Summer Interns:
They did a great job, and really understood SEO, as evident by these SEO Proverbs (copied below) they left on their last day, along with a sweet drawing…
Link Juice Proverbs
- Like most juices, its better when its not from concentrate (domain diversity in links)
- Your thirst for it can never be satiated (persistence and hard work pay off)
- Naturally squeezed link juice tastes better than bought (Two sides: One is light; one is dark – natural editorial links are the best)
- Link Free…Or Die Hard (Don’t be a juice hoarder, link to great external websites)
- Dirty linking is not the same as dirty dancing (meaning unclear…)
- I don’t always drink juice, but when I do, I prefer mR 7 link juice (not all links are created equal – spending 5x time getting one link might be more important than spending x time getting 10 links)
- The juice is worth the squeeze — outreach your content (Content doesn’t get links by itself. Do your content justice by promoting it to targeted parties)
It’s Not All Fun and Games
Just recently we got publicly criticized by SEO demigod Rank Fishkin for using homepage PageRank in a PR score (obviously I don’t use PR in my SEO measurements, but what SEOs would call old ways of thinking is still not flushed out of everyone at SeatGeek).
Well you know what Rand…
Oh and we also no-indexed every important page on our site, but we were able to make the best of that with a SEOmoz post that was promoted to the main blog. Our co-founder Jack weighed in on the no-index situation for this post:
The de-indexing was probably my most engrossing singular SEO experience at SG. Obviously it sucked a lot at the time, but it did have a cool positive takeaway in that (a) we didn’t suffer long-term damage and (b) served as material for a post on the seomoz blog. So maybe it was the best unintentional meta-seo play of all time.
Other SEO Takeaways
“Don’t take the measurements too literally — I’ve been surprised by the way that position shifts so much on a week to week basis. It’s tempting to think that we’re at fault, but such variance is natural. Also, adding to the fuzziness is the fact that the tracking tools aren’t always accurate.”
Not to say that measurement isn’t important, but in a data world where everything is black and white, it is important to get a little flexibility when reporting on SEO metrics. I am at the point now where something will drop from page 1 to 2, and unless I noticed a sitewide trend I will just address it and defer analysis for a few days to prevent wasted actions if this is just a Google dance.
Again from Russ:
“Importance of doing analysis before pursuing a strategy — It’s tempting to think that one can drastically improve SEO by rolling out tens of thousands of landing pages. Unfortunately, such a strategy can be actually deleterious to long-term SEO goals. I really have enjoyed how we use our blog and Adwords tools to test competitiveness / traffic before we build anything out. I’m not sure how commonplace this is among startups”
One of my favorite things to do is to use Adwords as a keyword research tool, not only for new opportunities, but to refine existing pages. This isn’t innovative, but I suspect people get lazy in this area (me not excluded at times…). Also, we have used our blog to test our ability to rank for different query types before committing dev resources to creating landing page, which is a great strategy when dev. resources are tight…plus it will earn you some respect from the dev team.
“SEO can be huge in Biz Dev deals and can entirely change the way we approach the economics of a deal.”
As mentioned above, some of our most respected links are from Biz Dev deals. Let the debate begin, but you can’t ignore this variable when assessing said deals.
Where We Are Going With SEO
Of course I am going to keep learning, stay on top of the trends as put out by the likes of Distilled (drink), SEOmoz, Wall, etc. and keep getting our pages on page 1 and beyond. But honestly that is just a short-sided view. As a company we don’t really succeed until my job doesn’t exist (in my opinion). We are a search engine. People should come to us for a far better ticket search experience than Google, or we should be incorporated better in major search engines…but so we dream. Until then I will keep doing what I do, and more with the help of you all. If I win this maybe all my friends and soon to be friends at Distilled can help.
Why We should win
This is a contest for startups. SeatGeek is a real startup, not some sort of marketing agency chop shop or hardly working social mobile local app. We have a revenue generating product and put SEO at the forefront of our acquisition strategy. I would safely argue that of all NYC startups <15 people, we are in the top percentile of SEO focus. SeatGeek is a company of 11 people where there is only one SEO person (not even 100% of time), but somehow 15 SEbrOs and ladies. We stretch our resources very effectively and we are really dope.
Note: Jalen Rose doesn’t work at SeatGeek
Thanks to those that helped me learn
I got a list here’s the order of my list that it’s in; it goes Rand, SEOmoz in general (especially WBFs), Art of SEO, SEO Warrior and the entire SEO blogging community.
And of course everyone that works or has contributed to SeatGeek. Oh, and if you think this is all we have done or got in us, take off your crown clown…