Yesterday at 9:12 am MST a major brand which runs its affiliate program on a major affiliate network sent out the following text to all of its affiliates:
We are scheduled to perform maintenance work on the [MerchantURL].com website and as such will be having an outage effective Monday August 8th from 10pm MST through Tuesday August 9th 6:00AM MST.
During this period we will not be able to take any orders through our website. We apologize for the inconvenience that this may cause. Please let us know if you have any questions.
Very seldom are the cases when merchants do something that shows they really care about affiliates. However sad this may be to admit, more often than not, it is the other way around — tracking goes down, websites go down… but merchants don’t say a word. While the affiliates of the above merchant would’ve benefited from a more advance notice (13 hours is cutting it really short), as some may be paying for the traffic they’re sending to the merchant, overall, I applaud the above step.
This incident also brings another possible scenario to mind. If it so unfortunately (for affiliates) happens that a website outage or tracking downtime occurs, and before you know it a period of no-commission time elapses, you as a merchant (or program manager) want to retroactively compensate affiliates the commission they would’ve earned had the tracking worked properly. I’ve had this happen in my affiliate program management practice before. Affiliates don’t care who’s fault it is, and appreciate it when you show you truly care.
HasOffers went down for about 8 hours from late last night until early this morning. Not that I drive any traffic through HasOffers networks as I don’t trust any company that uses them – but its an interesting case study based on the fact that HasOffers chose to put their source code into the ‘cloud’ and thus exposed their client base to every conceivable issue that all other sites on the Amazon EC2 may run up against. Not only are all HasOffers clients put at risk in a shared pool of other HasOffers clients – but they are at risk from every single company that hosts on Amazon EC2. A combined 15 hours of downtime in two months screams that this is not an effective mechanism for powering high volume affiliate programs and networks.
As an affiliate, not only do I not trust a single HasOffers network due to the zero barrier of entry and complete lack of financial investment in their business – I also feel as though the ‘cloud’ hosting environment presents a massive risk potential and up-time issues for any network dumb enough to trust their business on this cheaply built technology.
Hi Gary,
We did have some down time in the Virginia facility of Amazon EC2 yesterday, but it was not anywhere near 8 hours. In fact, it was approximately 10 minutes. Our fail over systems were engaged, but the service went back up before the switch. You can check amazon’s logs on this. This falls well within our SLA, but we are obviously very upset this happened. We are continuing to improve our global load balancing to decrease the amount of latency for failing over to other facilities in these events. As a final fail over, we still maintain our own servers in LA. So even if every Amazon instance went down, we would still be able to carry the load.
Would you mind emailing me a network name that claimed to be down for 8 hours? [email protected]
From the beginning, it has been our desire to put reliable tracking technology into everyone’s hands so that the entire industry may grow. There are always difficulties in attempting to move forward onto more advanced technologies, but we maintain that the future is in the cloud.
Peter, the downtime was quite significant and effected all HasOffers clients for much longer than ten minutes I assure you.
Check this blog post to see what your clients are saying about it:
http://www.wickedfire.com/affiliate-marketing/132889-wow-great-feature-hasoffers-thanks.html