Analytics are one of the most important parts of your marketing. They’re what allow you to see what you’re doing right, what you’re doing wrong, and what effect your changes have on your audience. Sometimes, something as simple as moving an image to the left or changing the position of a button can have a profound effect.
The most frustrating thing to deal with, regarding analytics at least, is the delay in reporting. Google Analytics is probably the number one most used analytics suite on the web today, because it’s robust and free, but there’s a delay in updating and reporting data. If you post an update on Facebook, you’ll see the data appear in Insights faster than you will through Google Analytics. This is all setting aside the issue of data sampling.
How much of a delay does Google Analytics have? Well, it will vary from site to site. Smaller sites – those with low traffic numbers and low numbers of pages – will report data fairly often, anywhere from a ten minute delay to an hour or two. Larger sites, with thousands of pages and hundreds of thousands of visitors, might take five or six hours. Extremely large sites, the Reddits and the Facebooks and all the other giants, those would have delays closer to 12 or so hours, though most large sites use either custom analytics scripts or solutions other than Google’s. That, or they use the premium enterprise version of Analytics, which gets priority data processing, but is designed for companies with tens of millions of daily visitors. Not something for the average Joe, that’s for sure.
Of note, it will also change based on the kind of information you’re trying to get. Basic statistics are easy to filter and report, so the data will update faster than intelligence reports and other deeper analysis of the data. The Visitor Flow report is on a much larger delay.
Now, I can understand delays with Google Analytics. Google may be a huge company with a virtually limitless supply of servers and computers doing their processing, but they also have a ton of literally global services. How much processing power do you think it takes to run search, and the algorithm? To receive and process the data from millions of web spiders? To handle and serve Google Drive, Google Apps, Gmail, and all their other services? To receive, analyze, and dispense data from Analytics? There are different degrees of infinity, and Google’s is close to capacity.
It’s a problem that any sufficiently large analytics company will have to face. The more users they have, and the larger the users they have, the more processing power they need to devote to monitoring and analysis. Most companies simply accept that there’s going to be a delay, and when data shows up faster, it’s a bonus for the user. However, some companies find ways to speed things up. They run locally, or they have sufficient power dedicated to the task that they can report data quickly and accurately.
What we’re talking about is real time analytics. When a user visits your site, you see data about them immediately, not ten minutes from now, not an hour from now, and certainly not a day from now.
The Skeptical Value of Real Time Analytics
One thing we have to question is whether or not real time analytics are really worth it. Sure, you get up-to-the-second data, but is that valuable? How can you utilize that data in a better way than you could if the data was on a 24-hour delay? There are, certainly, flaws with real time analytics.
For one thing, real time data means you’re swamped with more data, constantly shifting and updating. If you want to draw an analysis report, you need to snapshot the data all at once, or your report may use data from different times, and thus different numbers, adding fuzziness to your reporting. As it is, many marketers are already swamped with more data than they know how to use, and simply being awash in a sea of constantly-updating numbers is a novelty more than a valuable prospect.
Secondly, you have to keep resources in mind. Real time analytics take more work on the part of the server, sending and receiving data. This may slow down the performance of your site. It may be significant enough to hurt your SEO, though it does usually take a significant hit to cause that kind of an issue. There are resources beyond your servers, though, and it’s a drain on the resources of the people you have tasked to monitor your analytics as well.
A third possible issue is simply the viability of the analytics suite you choose. If you’re putting “real time data” ahead of specific reports, actionable intelligence, or actual data harvesting, you’re sacrificing tangible value for potential value. Imagine if you had a real time analytics service that did little more than simply count hits on your page. Compare that to something like a heatmap; the heatmap might be on a delay, but the data is very precise and much more useful. Though I will be fair here; most real time analytics suites today are capable of providing a lot of useful intelligence. This was more of a concern a few years ago, where “real time” was a buzzword used to sell sub-par tools.
There are two major concerns you should consider before you opt for a real time analytics solution. These are statistical significance and change agility.
Statistical significance is important because you have to consider sample sizes. Reporting on a half a dozen visitors isn’t valuable, it’s simply not enough traffic. Also, if you’re recording data on a minute-by-minute basis, you will see spikes and lulls in your traffic as they happen. Your data might even out over time, but if you’re seeing a lull and react as if your post is a failure, you’re not giving yourself time for the spike.
As for change agility, it’s the amount of time it takes you to react to your data. How useful is your real time data, if the process of changing something to react to it takes 48 hours? At that point, it doesn’t matter if the data is real time or on a day-long delay, you can’t react to it fast enough to tangible make immediate changes.
The Real Value of Real Time Analytics
That may sound like a lot of nay-saying, and to an extent it is, but I will say this; real time analytics solutions have improved a lot over the last few years, as processing power becomes easier and easier to obtain, data becomes easier to analyze quickly, and the whole analytics world becomes standardized. There are some real, tangible ways you can use real time analytics for the benefit of your site.
The first way you can benefit is simply technical; debugging issues with your site and your analytics itself. When data is reported on a day-long delay, it takes a while before you recognize an error. With real time analytics, you can see immediately if the numbers don’t look right. A broken tracking parameter, a faulty bit of code, a 404 link; these will all show up immediately and allow you to react quickly to fix the problem.
The second way is to see the immediate results of an action. For example, if you post an update on Facebook, you will be able to see the sort of initial flood of traffic, as well as any spikes from when it is shared. However, keep in mind that this is only useful for initial impressions. Tracking engagement or future actions based on this data is impossible.
The third way is with split testing. You can perform split tests in real time, making changes and seeing how the data reacts. However, in order to successfully pull this off, you need to have a large amount of incoming data. Remember statistical significance? If you’re getting thousands of hits an hour, then a split test can be pretty effective. Otherwise, you’re not getting enough data in a short enough time frame to make the real time reporting worthwhile.
And, of course, there’s the novelty factor. For those of us used to day-long delays on our analytics, seeing numbers change and increase by the minute will be extremely satisfying and interesting. However, that’s really just a pleasure factor, it’s nothing actionable or usable.
Real Time Analytics Options
But enough about all of that, right? I’m guessing a lot of you skipped right down to this section. You don’t want to know why or why not to bother with real time analytics, you want to know what options are available. Well, here are the best I’ve found.
Clicktale – Clicktale’s real time analytics are only “technically” real time, in the sense that the script fires once every 60 seconds while you use the free version, and once every 10 seconds if you pay for their advanced version. They record referral data, link tracking data, session information, and a handful of other metrics. They also have a session recorder, so you can play back what individual users are doing, including on-site keystrokes and mouse movements. However, data is only saved for 24 hours, or at max for 60 days if you have a premium account. This is very much not for long-term analytics use.
Gauges – One of the more interesting live analytics options, they have a very slick dashboard. The service starts at $6 per month and tracks up to 100,000 pageviews per month. Higher limits – and team sharing – come with larger plans. Their script doesn’t need to refresh your dashboard to show you new data, it simply updates automatically as new data comes in. That said, it’s not particularly deep information. They show you visits, referrers, search terms, geolocation, user agent and device data, and not a lot else. It’s the tradeoff at work; the data is real time, but it’s raw data, not processed reports. You have to make sense of the data yourself.
Woopra – This is one of the more robust options available, but with the caveat that a lot of the advanced features are tucked away out of sight of the casual marketer. You wouldn’t necessarily even discover a lot of the power of this analytics tracker without digging into the documentation. You’ll also have to have some understanding of analytics to create and read the reports you want to see. In other words, it’s going to take some setup time, but once you get things pinned to your dashboard, you have real time reports front and center. The kicker is the price; free up to 30,000 actions (pageviews, downloads, and other hits) per month, at which point the cheapest plan is $80 per month. The most advanced features, like custom reports, are enterprise-only as well.
Piwik – This analytics suite is notable for being free and open source. They have real time visitor tracking, including goal tracking for those visitor sessions. Their dashboard is customizable and can track multiple websites at once. Since they’re open source and very popular – nearly three million downloads and counting – they’re actively in development and adding new options all the time.
Clicky – One of the first and most well known real time analytics suites. All of the data they provide is real time, unlike many real time analytics suites, which claim to be real time but barely put more than a hit counter in the real time corner. They have heatmaps, on-site analytics, sophisticated bot filtering and referrer spam removal, and included uptime monitoring. The free version only tracks one site and only 3,000 daily pageviews, and doesn’t include most of the premium features. The cheapest paid version is $10 per month, up to 30,000 views per day, and comes with everything except heatmaps and uptime monitoring. To get those, you need the $15 monthly plan or higher.
Google Analytics – Yes, believe it or not, Google Analytics does provide some data in real time. However, don’t get too excited. The data they offer in real time is a small subset of the total analytics package. You can only see certain types of data and certain reports in real time.
The post How to Track Your Website Visitors in Real Time appeared first on Growtraffic Blog.