Search and Discovery writer
Sorry, there is no results for this query
Have you ever bid in an online auction by throwing in your highest amount at literally the last minute, only to have someone beat you out by microseconds by using sneaky bidding software? Can you imagine how disturbing it would be to trade a stock at the perfect time and then have the trading site’s software react so slowly that you lose a significant amount of money?
In this age of Google search, people take data-processing speed for granted, and, as in the scenarios above, it’s a serious problem if there’s a lag. This issue is just as relevant when it comes to search, as well as the more passive activity of information “discovery” through browsing on shopping and media web pages.
If you’re an online retailer, every single search is a potential conversion opportunity. If you’re an online shopper, every single search is important to you in terms of locating what you need and not wasting time. If you can’t quickly find what you want, you’re not going to stick around. You want to find what you want now, in real time. OK, well if not in real time, then in near real time.
When it comes to Big Data, near real time is a loosely defined term, as what’s considered “near” varies with applications and websites.
According to the Free Dictionary, near real time (NRT) pertains “to the timeliness of data or information which has been delayed by the time required for electronic communication and automatic data processing. This implies that there are no significant delays.”
Hmm; sounds kind of like real-time processing, doesn’t it? Let’s get more into the nitty-gritty. There are three types of data processing:
You know that real-time processing is the fastest type, typically used when information must absolutely be made available immediately or else, such as when a broker is trading stock.
Batch processing is a more economical option that tolerates some latency. Batches of content in a system, for instance, banking transactions or software updates, are saved up and then processed all at once, but less frequently than might be desirable if time were truly of the essence. Batch processing may take a while (even days) to be completed, but it’s a reliable, widely used option in the business world.
That leaves near real-time processing, the middle ground between these fast and slow processes. NRT is for when there’s a need for some speed, but when relatively moderate, “acceptable” speed is more than adequate.
In terms of search, NRT refers to indexing. For instance, documents in an educational institution database could be made available to be retrieved by students searching “almost immediately” after they’ve been indexed. “Almost immediately” could mean a timeframe of minutes, seconds, or milliseconds.
Of course, you’re probably thinking, that’s not very good. When it comes to searching, real time would seem so much better. A real-time search engine would be the best, so why would you want anything else?
It’s true that near real time is technically substandard by comparison; it means you’re not optimizing to provide updates as fast as possible. But before you decide that’s unacceptable, let’s keep things in perspective here and reflect on the fact that when it comes to human beings using software, near real time is still super fast.
Plus, as with anything, there are some drawbacks to vaunted real-time processing, such as the need to set up costly high-performance hardware and difficulty with auditing the data.
OK, so we’re probably in agreement now that if you’ve got a website that has a search feature, near-real-time results would be a pretty good thing to provide to your users or customers. But in case you’re still partial to the concept of real time and not quite convinced of the beauty of near real time, let us count the ways that near-real-time search is something to celebrate.
Having modern search functionality on your website that dynamically updates web pages as your catalog content is updated translates into a better user experience that could include:
Here’s a theoretical online-retail use case: you’re going camping in a couple of weeks, so you’re shopping for a basic tent. You go to an outdoor supply store website homepage, enter your tent search terms (“basic camping tent”), and browse the many, many options.
The site also recommends that you check out some fancier types of tent (like ones that are geodesic-dome shaped) to discover as you’re perusing the no-frills ones.
You settle on a basic tent that seems like the best quality for the best price. “Done,” you say, imagining waking up in the wilderness in that tent.
But wait, where’s the Buy button?
On the right side of the product detail page, you notice the message: “Hold on camper, give us a minute while we see if this tent’s in stock.” (Well maybe not in those words, but you get the drift.)
Or worse, the message says that your dream tent indeed is not in stock. (Actually, a shipment has just come in, but the merchandising team hasn’t gotten around to updating the listing.)
Now you have to either go car camping or start your tent search over. Grr.
What if this website had simply offered fast, efficient search? This camper could have quickly nabbed their preferred tent and headed out.
It’s safe to say that a personalized online shopping or browsing experience is something most people now expect. Shoppers want to be catered to, and they understandably don’t want to waste time. It’s not OK for your site to give them the equivalent of a ticking clock along the lines of “Hold on! We’re preparing your personalized content.”
And if your business’s search engine isn’t operating at near-real-time speeds and gets bogged down by personalization tasks, that’s a significant problem, of course. Your users will probably give up and jump off of the site, perhaps never to return. Not good.
Let’s say you’re running some type of news site. You’re probably concerned about your search query processing times, and rightly so, as when it comes to news, timeliness is a must. Media consumers who are used to frequenting sites like Google News appreciate knowing the latest update shortly after reporters hit Publish. And if your “trending topics” are even slightly out of date, your users, accustomed to hitting “Refresh” every few minutes to get the latest scoop, aren’t going to develop much trust in your platform.
Customer service is another place where data-processing time is of the essence. If you’re an unhappy customer who calls customer support, you want them to be able to pull up all of your data instantly. What you don’t want is the CS rep asking you to provide details that you’ve already entered on your phone keypad or online, but that apparently haven’t been processed and stored in the system yet or been allowed to be transferred between information silos. You don’t want to be put on hold for a dumb reason.
Near-real-time-information support leads to speedy problem resolution.
Data processing needs to be able to keep up with the business it supports. To facilitate such data agility, a company needs a competent search tool that can give people access to the latest available searchable data they’re seeking. For example, on an ecommerce site, you’d want your inventory to be absolutely current so that merchandisers can access critical details in order to plan strategy and engage in productive decision making.
When it comes to business intelligence and operational success, processing speed is truly of the essence.
Now that you’re enlightened about the nuances of the search-data-update process, you may be thinking about how you can upgrade to near-real-time content search on your website. Why not give your users the information they need without delay, and, as a result, make them happy while also driving positive results for your bottom line?
Algolia’s near-real-time analytics API is one proven option. Our search engine uses advanced machine-learning algorithms and indexes data in near real time, providing your users with speedy search results that lead them straight to the right content on your site or in your app. They can then get on with their day (and, if applicable, leave you positive reviews).
In terms of Algolia’s response times, how fast is fast? There’s no delay between when a searcher types a letter in the search box and when they get a search results page to peruse. We ensure search availability 99.99% of the time. That’s nearly as fast as…well, not that, but it’s fast.
If you’re ready to find out how we can start speeding up your site or app experience with near-real-time search, we hope you’ll connect with our team. Are you a developer? You can start building out your site’s new search solution for free. Check out the ways Algolia can help your business grow today.
Powered by Algolia Recommend