2344 0
An image of a team discussing integration of API

On-demand data stream

The proliferation of “as a Service” platforms is a testament to how the cloud ecosystem has changed (for the better) the way in which businesses can more cost effectively and dynamically consume services. One of these aaS’s is DaaS, or Data as a Service. Essentially, DaaS allows businesses around the world to access quality, secure, on-demand, real-time data streams without the constraints of traditional, siloed datastores. Put simply, businesses can focus more on data driven actions and less on admin.

Accessing new data sources with DaaS

It’s impossible for businesses to have access to all the data they require based on their internal sources alone, especially when it’s not their core capability (one of the main reasons any ‘as a service’ is used). There is the ability for businesses to harvest or capture new data from disparate sources, but it quickly becomes an issue with regards to quality, consistency, maintenance and even access rights. DaaS removes these issues by transferring these hassles to a specialised DaaS provider like PredictHQ, that gives you the agility to integrate quality data into your products via a cloud service.

By breaking down the barriers around data integration, businesses can focus on generating new insights or developing features/products which previously would have been less cost effective to build.

Speeding up integration

The initial challenge that comes with all products that require some form of integration is speed. If DaaS-based services can reduce the friction/barriers to integrate then it naturally improves the speed at which the product can be assessed and integrated. The areas where we focus our speed to market efforts are relatively straightforward (outside of your standard reliability measures):

  1. Documentation – we keep our API documentation up to date and as clean and simple to digest as possible. This will evolve as we make our documentation even more informative with example-based, bite-size video tutorials.
  2. SDK (Software Development Kit) – if we can reduce the amount of code a software engineer has to produce within their chosen programming language to leverage our API then it’s a no brainer. We’ve already created SDKs for Python and JavaScript with many more planned in the coming weeks (not years).
  3. Feedback – we simply listen to what our development community is telling us and if appropriate, implement the feedback as soon as possible. Feedback from the coalface is something all businesses should nurture regardless of the platform or service.

It’s not just about the data

Whilst many DaaS services focus on providing high quality, on-demand data some take it that step further like Mapbox (who we use and love) that not only provide their mapping service but also useful tools like geocoding or map styling.

In the case of PredictHQ we knew that providing a single stream of aggregated and standardised global event data via an API is hugely powerful. However, we wanted to take it a few steps further and provide additional services within our event intelligence API.

1 – Event Ranking

When we pull in disparate data from around the world like concerts, festivals, school holidays, public holidays, observances, conferences, sports, severe weather, disasters and many more, it’s a huge undertaking to de-duplicate and standardise. But even after doing all that, the data is still relatively one dimensional, which is why we set about enriching every piece of event data we consume with complementary data such as venue information, artists’ social profiles, location population, etc. By doing this for the millions of events we process, we created a proprietary ranking attribute ranging between 0 – 100 (from minor to major) to represent the importance of each event. This rank can be used along with location, category, labels, date, and more by businesses so they can focus on what they need.

2 – Event Signals

Even with ranking taking you most of the way there, with the sheer volume of data available via our API, it’s difficult to distinguish the signal from the noise. So we set about evolving a feature within the API called Event Signals, which in its simplest form helps finding what events impact your business the most by location. Event Signals allows businesses to discover and focus on what events matter to their business to make more informed decisions.

3 – Extra Services

It’s cool to be able to search events relevant to your business, but sometimes you just want to be told (or your application to be told) what to do. We allow you to set up search criteria via the API, and define where we should send new events matching these criteria.

For example, you’re selling a holiday package for a family trip to New York, but are worried that severe weather will lower demand, while you keep your ad spend high. With PredictHQ, you can create a search for severe weather in New York. If an instance matching your search criteria occurs, we can send a notification to your Ad platform and/or send an email to your team, to pause your ads.

Live example – context to create urgency and improve conversion

If you would like to see a specific example of a product that has been created from our Data as a Service, then you can test out our contextual booking alert. This simple example takes your standard booking alert found on many travel based eCommerce websites and drives urgency through better context or answers the “why” around the lack of availability and increase in pricing. Alternatively the messaging can also be used to encourage customers who are looking to book early to transact before prices are likely to increase and availability decreases.

We have a number of customers such as Airportrentals.com who have customised this simple implementation within their result pages to materially improve conversion by up to 22%.

booking alert default image

In this article

Join the Conversation