API Ratings August 17-24, 2020: Cisco Spark

The time of the week has once more come around for API ratings from the last 7 days! We now find ourselves heading towards the end of August and we are experiencing torrential rain here in my part of Europe today.

We launched API.expert earlier in the year as a simple way to provide everybody with insights into the API economy and the leading APIs in different categories in particular. The basic service is free and will remain free, but other, related services will be built on top of it, including your very own API.expert dashboards if you want them!

If something is missing that you think should be here, don’t hesitate to drop us a
line!

Now, onto what the past week tells us about the state of the world of APIs as we dive
into the third full week of August 2020.

API Ratings: The Headlines

We look at over 200 API ratings, but pull all the metrics together to give you a general feeling for the service quality for an organization’s APIs in a particular category.

We do see that certain providers consistently vie for top spot in their category including GitHub and Google with others doing less well such as Halifax and the NHS.

Top API Ratings – CASC Score

Week Ending 24 August 2020

A stable week for once this week:

• COVID-19 Finnhub API swapping with ECDC (European Centre for Disease Prevention and Control) in Covid-19
• BitMEX exchanging places with Coinpaprika again in Cryptocurrency Exchanges

Nordea Bank retains the crown once again for the private sector with an exceptionally good CASC score of 9.92 (up from 9.89 last week). A CASC score of over 9.00 is very good and one of 9.50 or more exceptional.

Five of the nine categories are headed by an organization with a CASC score of 9.50 or more this week down from six last week (but one Google has a CASC score of 9.49 in Search). Sustaining a CASC score of >9.25 over a period of several weeks is a good showing and congratulations to those organizations that achieved it.

Top API Ratings – Latency

Week Ending 24 August 2020

Yet another stable week this week in this stable category with no changes again to the top performers.

Google takes the overall title again with a median latency of 60 ms in Social Networks, up from 57 ms the previous week.

An important caveat: medians can be misleading! An API might have a fast median latency but produce many slow outliers.

These won’t affect the median, but they mean that users can experience many calls that were unacceptably slow. Just being fast isn’t everything. You have to be reliable too if you want to have good APIs and get a high CASC score!

FTX tops the Cryptocurrency Exchange category by median latency. This week they are up down to fifth out of 18 overall with a CASC score of 8.70, placing them solidly in the Green Zone. Let’s hope they can keep up the good work.

But it is still worth remembering that it’s no good just being fast if you are flaky, although being fast does help!

Worst API Ratings

Week Ending 24 August 2020

A fairly volatile week this week with four changes: Cisco Spark a new entry in Corporate
Infrastructure, Open Bank Project a new entry in PSD2 Banks and Foursquare a
new entrant in Search and nobody below 8.00 in US Government.

Coronavirus DataAPI (thevirustracker.com) takes the overall wooden spoon with a poor CASC score of 4.32, placing it very firmly in the Red Zone.

Something of interest

Cisco Spark has not been having a good week and has dipped into the Amber Zone. Let’s have a deeper look at what is going on. We monitor six Cisco Spark (WebEx Terms APIs), four of which had CASC scores of over 9.00 and the fifth a CASC score of 8.92. So that’s a pretty good effort there and if weren’t for the sixth!

Cisco Spark List Rooms is the cause of Cisco Spark’s bad week. It has a CASC score of just 1.35 (ouch! That’s not a CASC score you see in the wild too often). And it is so bad because it is performing so badly on all four metrics.

They have a pass rate of just 82.09%, which corresponds to more than 30 hours of downtime over a week. A particular bad day was 18 August when most of the failures occurred, but this wasn’t a simple outage as there were still some passes sprinkled in within the 502s. But the endpoint is also incredibly slow with a mean latency of 15.3 seconds. Which is one hundred times what you hope for a good endpoint (it does vary depending on what the endpoint does, of course). We can see that this is from all clouds and locations, so it is definitely the backend that is causing the problems. The percentage of outliers has actually gone down since last week. It is now just 40.0% Unsurprisingly, the standard deviation of the latency is also high.

The endpoint hasn’t returned a failure since 19 August, so there is a good chance that the CASC score will be at least somewhat higher next week since the pass rate will be higher. Nevertheless, this is whatever way you look at it a very slow endpoint. Now, it might that the operation to read the list of rooms is for some reason intrinsically slow, but would still pay to investigate exactly what is going. That’s why you need to actively monitor your APIs. Most of Cisco Spark’s endpoints are already in tip-top condition. But List Rooms isn’t and there probably isn’t a good reason for it. Don’t assume that all your endpoints are good just because you think most of them are. Make sure they are all working properly. Make sure you know what all your APIs are doing right now.

See you again in a week as we look once more at the State of the APIs as discuss the last week of August. Also don’t forget to look out next week for our August monthly report.

Do more with monitoring!

If you are new to APImetrics, sign up for a free trial so see how they could work for you with our free API set.

If you have any questions, please don’t hesitate to reach out.

If you’re new to API monitoring, you can learn about our comprehensive features for testing, monitoring, security and performance and sign up for a free trial account.

CHAT TO US ABOUT YOUR OWN RATING REPORT

OR

DO IT YOURSELF WITH A FREE TRIAL