Should Brightcove Build Its Own CDN

Brightcove, the leading OVP, has reached $100M in sales in 2013. Their economy of scale, impacting overall cost reduction, is improving. As the volume of digital media delivery increases (measured in Gbps), CDN costs drop. Brightcove has partnered with Akamai for its CDN services. Some industry pundits have noted that Brightcove’s CDN costs are significant. The question is should Brightcove build its own CDN?

The Financials
According to Brightcove’s 10-Q ending 9/30/13, it booked revenues of $28.5M, with a Cost of Revenue (COR) of $9.2M, and 6,374 customers in 70 countries. If we break down the $9.2M quarterly cost, into a monthly amount, the COR is $3M/monthly. This includes costs for hosting, professional services, support, CDN, benefits, salaries, data centers, overhead, etc. How much of that $3M/mo cost of revenue is associated with CDN services, is it $250k/mo, $500k/mo, or $1M/mo? I have no idea, but let’s guesstimate at $300k/month.

Monthly CDN Costs
If they spend $300k/month on CDN services, they should build their own mini-CDN, similar to what Netflix has done. The $300k/mo equates to $3.6M yearly. Brightcove can reserve $600k annually for Akamai, and roll out its own CDN for $2M, saving it $1M per year. If Brightcove spends more on CDN services like $6M per year, it will save a few million. However, it shouldn’t rely solely on their own CDN for media delivery.

Brightcove can use their own CDN, in conjunction with Akamai’s for media delivery. I would split the traffic, 80% – 20%, where Brightcove delivers the majority of it. If there’s any hitches in Brightcove CDN platform, they can switch over to Akamai during the downtime. In Brightcove’s case, building and deploying a CDN is not as much as one would expect. I will cover that in my next post “The Cost of Building a CDN”.