Online marketing for startups
What to do on your first week to get started.
Creating your startup’s online identity is an often overlooked early step. Many startup founders (us included) are reluctant to unveil anything about the company out of shame that the product is not finished.
Yes, shame. Loveseat (a mobile market to buy used furniture) is still a month or so off from launch. We want the world to see our vision, which is not what we have today. Because of this, our instinct was to wait until we were ready for launch to unveil anything.
We were chatting with our advisor, Gabe Weinberg of Duck Duck Go, about our plans for gaining traction (an outside perspective is enlightening). He questioned why we were waiting for launch and ultimately gave us the push we needed to get something out there.
Taking the “we’re doing this for testing purposes” approach to online marketing made the whole exercise easier to swallow. Also, we only spent about 8 hours total on this, so it didn’t take too much time away from some of the more proprietary stuff we are working on.
Disclaimer: This post covers the most basic topics for online marketing / customer acquisition analysis. It’s a very “hacky” approach — which I would argue is exactly how a startup should do things.
Why bother with your online identity when you are pre-launch?
- It’s a tool to get the word out to early fans & you get to build on it.
- It’s a platform to test customer acquisition costs. Learn what works.
Establishing an online brand identity:
———————————————————————————————
If you are like most founders, you have extremely high aspirations for what your brand should look like. You want the perfect url, perfect logo, perfect colors — everything. You may even be tempted to hire a design agency, do a fancy video, etc.
Don’t do this. No one is going to remember what you looked like your first week. Just get something out there to start testing and building a following. We did all “design” ourselves by taking our mobile splash page and modifying it in Photoshop…nothing tricky.
Things we did this week:
Launchrock Page:
http://signup.loveseatapp.com/
Time to build out: 1 hour
Thoughts: We needed a landing page that would also act as our temporary homepage. For us it was really important that we had a form to collect email and zip code.
The Launchrock page got almost 100 signups in under a week. This is pretty good. We even signed up a bunch of cool vintage furniture stores and importers via our marketing efforts. Perfect.
Building out a Launchrock page is dead easy. If you are looking to get something up quickly, it’s a no-brainer. It has two useful features I like:
- Trackable URLs: Although they hide this feature as user referrals, you can use those trackable links to see the effectivness of online advertising.
- Google Analytics: You can plug in your Google Analytics property id and start collecting basic analytics instantly.
It gets tricky (maybe impossible) when trying to do sophisticated things like Google Adwords tracking or Analytics Goal Tracking. We were able to get some things going by modifying Launchrocks’ javascript (instruction here), but it’s a really ugly work-around. We’ll very likely end up building our own landing page w/ form in Django so we can do more advanced stuff before our launch.
Facebook Page:
https://www.facebook.com/loveseatapp
Time to build out: 5 minutes
Thoughts: Right now its only purpose is so we can run Facebook ads properly. We have a whopping 13 fans.
Twitter Profile:
https://twitter.com/loveseatapp
This also took about 5 minutes. Right now it’s only purpose is so we can run Twitter ads properly. Yes, 4 followers.
Google+ Profile:
https://plus.google.com/b/106055749303466121036/106055749303466121036/posts
This also took about 5 minutes. We will be using it to test content marketing via Google+ and see how it compares to other content platforms (Medium/Tumblr/Wordpress).
Testing customer acquisition costs:———————————————————————————
This is the part I really care about. The inner scheming geek in me LOVES this stuff. I want to know two things:
- How much does it cost us to get a new user via a channel?
- Is the supply of new customers we can get via this channel limited?
This week we wanted to learn about customer acquisition costs on Adwords, Facebook, & Twitter. What we learned was pretty surprising.
Google Adwords:
Total spend: $63.71
Cost-per-click: $1.72
Good old tried and true Google had the highest cost-per-click (and therefore highest acquisition costs) of any channel we tested. We targeted users in San Diego via the search network only — aka, no adsense ads.
The concentration of mobile users via Google adwords was much lower than what we saw on other channels, so I think it’s safe to assume this won’t be a major channel for us.
Disclaimer: Our ads sucked, the copy was meh, the landing page had a really low quality score — so I can’t really hold this entirely against Google.
Facebook ads:
Total spend: $6.97
Cost-per-click: $0.35
Facebook ads are ridiculously easy. We limited our ads to San Diego residents over 18 years old targeted based on broad categories of “recently moved” & “home and garden”.
You simply can’t do that with Google. Also, Facebook ads format perfectly in a mobile feed. Next up we’ll run ads based on more specific interests.
The big takeaway is that Facebook ads have a dramatically lower CPC than Google for our purposes. Our assumption is that there will be enough users to target via this channel over time.
Twitter ads:
Total spend: $28.86
Cost-per-click: $0.72
Twitter ads are also really interesting. They seem to be a better fit for us given the near real-time nature of the targeting. You can target the ads based on twitter search (less frequent matches) or feed content (more frequent). In other words, you can show an ad when someone uses Twitter search or you can promote an ad in their feed if they tweeted recently with triggering keywords.
We ran our ads to San Diego users only targeting keywords “Craigslist” “furniture” and “ikea”. The results were pretty strong. With some better targeting we expect these to improve and the CPC to come down. Also, since these ads are based on Twitter updates, we don’t think there will be a limit in terms of potential audience for us.
Twitter lead-gen cards:
Total spend: $49.82
Cost-per-click: $0.43
Twitter lead-gen cards are my favorite thing. Ever. This is displayed to Twitter users in their feed as a promoted tweet targeted just like a Twitter ad. The difference is when the user clicks the button on the card that says “Join the beta list”, Twitter adds the user’s registered email address to a csv file they send to us.
What this means is that a CPC is really a CPA. We have now “acquired” the user’s email address. While they aren’t yet a user of our platform, we can reach out to them to explain the benefits of Loveseat.
Just for kicks, we did a second campaign w/ Twitter lead-cards targeted at startup people in San Diego (keywords: startup, techcrunch, vc, entrepreneur, github, etc.). Guess what? These folks converted at a whopping 3.41% which is 50% higher than the other target group. Go nerds!!
So, we spent a total of $149.36 + 8 hours of time to learn some key information. What did we learn?
We learned that our rough acquisition costs for a user ranges from $1.95 to $11.18. This is before doing any optimization and with some not-so-carefully thought through ads. We also learned that Twitter & Facebook are probably a better fit for us to promote our mobile platform.
The main takeaway is there is some opportunity here. And, we can talk to someone (like a VC if we were raising) about user acquisition and have some data to back up our claims.
Equally importantly, we now have accounts set up on a bunch of platforms we will use into the future.
And finally, most importantly, we have nearly 100 people who are eager to learn more about Loveseat.
There are people out there who will say: “Yea, I’ve heard of that”.
=)
Final notes:
We were only able to approximate CPA due to some limitation with Launchrock (not a knock against them). However, we know that our landing page has a 15.45% conversion rate. So, we can take the cost-per-click spend above and multiply by ~6.5 to get the CPA. Remember, the key at this stage is directional information.
One tool that was mentioned here is Google Webmaster Tools. It’s used for a number of things in conjunction with Google Analytics & Adwords. It didn’t help us just yet given the setup of our landing page, but you should definitely get on it.
