Brandwatch, Now Faster Than Ever
By Phillip Agnew on March 15th 2017Read this article on our full site
For every second faster Walmart has been able to make pages load, it sees a 2% lift in conversions. Time is money - so we've just made Brandwatch faster.
When asked what his dream product would do, Apple co-founder Steve Wozniak answered: “give me more time”.
He didn’t want to live forever. Instead, he wanted to decrease the amount of time he spent doing non-critical tasks.
You’re probably thinking ‘he’s a billionaire, of course, he wants time, it’s the only thing his money can’t buy!’
Well, you’re right. But the thirst for speed affects us all.
Google discovered that even a 400-millisecond delay could result in 8 million fewer searches per day.
And for every second faster Walmart has been able to make its pages load, it has had a 2 percent lift in conversions. (CMS Wire, 2016).
It’s clear that everyone values their time.
What we’re working on
Speed is something we’ve focused on at Brandwatch for the last 16 months. We don’t want our clients spending hours trying to collect insights. We want the process to be quick and seamless.
Go in, search for data, uncover an insight or an answer. Leave and action that insight across your business. Our goal is that you spend more time actioning an insight (improving a campaign or tweaking a product) and spend less time finding answers.
That’s simple in principle, but hard in practice, especially when working with large data sets.
Here’s what we’ve done.
Up to 50x speed improvements when charting large data sets
Some of our customers collect millions of posts, tweets, comments and mentions each day. They use Brandwatch to group all these and conduct top-level analysis on the whole data set.
As you can imagine this can be time-consuming. When we first created Brandwatch this sort of analysis wasn’t possible. Today it’s possible, but still extremely difficult.
In fact, most of our competitors crash when performing these tasks. To cope, they default to sampling the data meaning clients only collect a portion of the total data set.
We challenged ourselves to go one better. Analyze full 100% data (without sampling) at speed. Early last year we implemented these improvements and we’re now seeing genuine improvements.
The platform is now up to 50x faster when charting large data sets.
Faster Mentions and Search component
Our mentions and search component reveals all the individual posts, comments, tweets etc. behind a data set. Pretty simple.
However, when customers used this component when analyzing a lot of data, it started to become slow and cumbersome.
Today we have enabled improved functionality for all our clients.
This update will:
- Significantly speed up search-within mentions in the Mentions & Search component
- Speed up click-throughs on any component in Analytics, including Topics.
Our customers often tell us that the ability to reveal key topics behind a conversation is extremely valuable. Being able to do this at speed will seriously decrease the time it takes to find an insight in your data.
Here’s an example. We’re looking at over 7,000 mentions of people saying the love they latest IOS update. It takes just seconds to load the individual mentions and the overarching key topics.
Okay, so this doesn’t directly relate to speed, but it’s still important. If the platform goes down you’ll lose all your work and you’ll end up spending a lot of time redoing the analysis.
Or, even worse, your monitoring a mega-event like the Superbowl or the Olympics and the platform stops collecting data. You’ll miss crucial intel on your brand’s most important event.
We’ve introduced some crucial changes over the last year to make sure our platform is mega-event ready. This included buying in more hardware to increase our capacity and optimizing our existing hardware.
In fact, during the last two Superbowls, Brandwatch was one of the only social listening platforms that stayed fully stable despite collecting hundreds of millions of mentions.
Our product is fast, which is great. We think it’s industry-leading in that sense, which is even better. But what we’re really happy about is what that will let our clients do:
- Answer those awkward 4:50 pm boss emails in minutes, not hours.
- Have the time to enjoy a proper lunch break.
- Spend your morning at a yoga class – not loading up a data set.
- Go out for coffee, rather than sipping instant while waiting for data to load.
These are just suggestions. Spend your time as you wish, and we’ll keep working to make sure you have as much spare time as possible.