Tips on Facebook ad campaigns February 16, 2011Posted by jeremyliew in advertising, facebook.
A couple of weeks ago Webtrends analyzed 11,529 Facebook ad campaigns representing 4.5bn impressions to see what conclusions they could draw. It’s worth reading their short white paper on Facebook Advertising Performance Benchmarks and Analysis. Some highlights include:
- Click through rates steadily increase with age up to 65
- State by state CTRs are relatively flat with the exception of North Dakota and Wyoming being substantially higher than the other states
- Fun/entertainment focused campaigns work better than other types of topics
- People who didn’t attend college click more than people who did
- People who attended college use their friends as a filter to determine who to click on more than people who didn’t attend college
- FB ads need new creative after three to five days
I’d urge you to read the whole thing. It’s only seven pages long.
Today is a good day to market to cheating spouses February 15, 2011Posted by jeremyliew in apps, Ecommerce, seasonality.
Bloomberg Businessweek has an interesting article on the infidelity economy this week, noting that registrations for AshleyMadison, the online dating site aimed at married people, spike the day after Valentine’s Day. But nowhere near as much as they spike the day after Mother’s Day and Father’s Day.
It’s an interesting example of the sometimes ideosyncratic seasonality of web businesses. Knowing the seasonality of your business can help you market to, merchandise for and communicate with your customers most effectively.
Everyone is familiar with the Q4 seasonality of retail businesses, driven by the holidays. But each industry has its own annual seasonality cycles that may be less obvious in foresight, but are always obvious in hindsight. For example, in Shoedazzle’s first year of operation, we missed a few months of the winter boot season. Having the wrong merchandising mix from October to January in 2009/10 definitely dampened sales – women don’t buy as many open toed shoes when there is snow on the ground!
Another example is Mercantila, a company that sells a lot of exercise equipment, rowing machines, elliptical trainers and the like. They see sales spike not in Q4, but in Q1, when people are making their new years resolutions. When many online retailers are easing up on their online marketing and SEM, Mercantila is ramping up what it is willing to spend to reach new potential customers.
Smartphone app developers have learnt that the launch of new iPhones and heavily marketed Android phones is a period when they can rapidly increase installs, as are the days after Christmas, when many new phones are getting unboxed. The first thing a new smartphone owner does, is go to the appstore to get some apps. Being high in the “best-sellers” lists at that time can provide a real boost to your installs. Making sure that your release at that time is bug free, well reviewed and fully featured, is of course important too.
Seasonality is not confined to annual cycles either. As the number of ecommerce businesses relying on email increases, rules of thumb for weekly and daily seasonality are also starting to develop. For example, Mondays and Tuesdays of the first and last week of a month are when people tend to balance their checkbooks, so are bad days for subscription services to send e-mails. They have a higher probability of getting canceled on those days and don’t want to draw attention to themselves. On the other hand, emails sent Sundays after Church show high open rates for some ecommerce merchants. And for deal sites, open rates are higher when you are above the fold when someone first opens their email, which rewards sending emails in the weekday 5am-7am time slot to work based email addresses.
We continue to be very enthusiastic about the tremendous amount of opportunity in the Enterprise Infrastructure sector for 2011. In the past few years, we’ve seen significant innovation in technologies such as virtualization, flash memory and distributed databases and applications. When combined with business model shifts (cloud computing) and strong macroeconomic forces (reduced R&D budgets), a “perfect storm” is created where the IT ecosystem becomes ripe for disruption. Startups can take advantage of the changing seas and ride the subsequent waves to emerge as leaders in new categories. For this post, I’ll highlight three categories where I believe we’ll see significant enterprise adoption in 2011 – big data solutions, use cases for cloud and virtualizing the network. Startups in these categories are now at the point where ideas have become stable products and science experiments have transformed into solutions.
1. BIG DATA SOLUTIONS GROW UP
There’s been a lot of “big noise” about “Big Data” for the past couple of years but, there has been “little” clarity for the traditional Enterprise customer. Hadoop, Map Reduce, Cassandra, NoSQL – all interesting ideas, but what Enterprise IT needs is solutions. Solutions come when there are products optimized to solve the challenges with specific applications. Most of the exciting, fast growing technology companies we hear about daily (Facebook, Zynga, Twitter, Groupon, LinkedIn, Google, etc) are incredibly efficient data-centric businesses. These companies collect, analyze and leverage massive amounts of data and use it as a fundamental competitive weapon. In terms of really working with “Big Data,” Google started it. Larry and Serge taught the world that analyzing more information generates better results than any algorithm. These high-profile web companies created technologies to solve problems other companies had not faced before. In this copycat world we live in, Enterprise IT is ready to follow the consumer-tech leaders. The best enterprise companies are working hard to leverage vast amounts of data in order to make better decisions and deliver better products. At Lightspeed, we invested in companies like DataStax (www.datastax.com) and MapR Technologies (www.maprtech.com) because these are startups building solutions that enable Enterprise IT to work with promising Big Data platforms like Cassandra and Hadoop. With enterprise-grade solutions now available, I expect 2011 to be a year when tinkering leaps to full-scale engagement because these new platforms will deliver a meaningful advantage to Enterprise customers.
2. CLOUD COMPUTING FINDS ITS ENTERPRISE USE CASES
The hype around “Cloud Computing” is officially everywhere. My mom, who is in her sixties (sorry Mom) and just learned to text, recently asked me about Cloud Computing. Apparently she’s seen the commercials. In Enterprise IT circles and VC offices, there’s a lot of discussion around “Public” clouds vs. “Private” clouds; Infrastructure as a Service vs. Platforms as a Service; and the pros and cons of each. It’s all valuable theoretical debate, but people need to focus on the use cases and the specific economics of a particular “cloud” or platform configuration. As of right now, not every Enterprise IT use case fits the cloud model. In fact, most don’t. But there are three in particular that definitely do — application management, network and systems management and tier 2 and 3 storage. At Lightspeed, we’ve invested in a number of companies such as AppDynamics (www.appdynamics.com) and Cirtas (www.cirtas.com) which deliver solutions that are designed from the ground up to enable enterprise class customers to leverage the fundamental advantages of “Cloud Computing” – agility, leveraged resources, and a flexible cost model. Highly dynamic, distributed applications are being developed at an accelerating rate and represent an ideal use case for cloud environments when coupled with a solution like the one offered by AppDynamics which drives resource utilization based on application level demands. Similarly, Enterprise IT storage buyers have gotten smarter about tiering data among various levels of storage media, and infrequently accessed data is a great fit for cloud storage. Cloud controllers like the one offered by Cirtas enable enterprises to have the performance, security and reliability they are used to with traditional internal solutions but leverage the economics of the cloud.
3. VIRTUALIZING THE NETWORK
To date, the story of virtualization has been primarily about servers and storage. Tremendous innovation from VMware led the way for an entirely new set of companies to emerge in the data center infrastructure ecosystem. At Lightspeed, we talk about the fundamental pillars of the data center as application and systems management, servers, storage, and networking. Given all the advancement and activity around the first three, I think it’s about time the network caught up. As Enterprise IT continues to virtualize more of the data center and adopts cloud computing models (public or private), the network fundamentals are being forced to evolve as well. Networking solutions that decouple hardware from software are better aligned with the data center of the future. Companies such as Embrane (www.embrane.com) and Nicira Networks (www.nicira.com) are tackling this challenge head on and I believe 2011 will be the year where this fundamental segment of data center infrastructure starts to see meaningful momentum.