Build your own

Add your own data service or custom engine to Pipeline API, or derive new properties from the ones it already returns with a short script. Combine the result with our device detection, IP intelligence, and geolocation data in one response.

An example element, a date of birth in and a star sign out

{
  "dateofbirth": "1992-11-05",
  "starsign": "Scorpio"
}

Example A custom flow element that turns a date of birth into a star sign

What you need to know

🔗

A pipeline is a chain of elements. The evidence from a request, such as its headers and IP address, goes in at one end. Each element adds its properties, and an element can read the properties the elements before it added.

Evidence, the request's headers and IP address go in. Our engines add device and location properties. Your element adds your own data. Derived properties add new properties worked out from those. One result, your application reads every property from one place. Evidence headers, IP Our engines device, location Your element your own data Derived new properties One result every property

Your application reads every property from one result, whichever element added it.

Read about Pipeline API

🔗

This example shows a very simple flow element which takes a date of birth and returns a star sign. Although a basic (and not all that useful) flow element, the example demonstrates how you can start to implement your own flow elements.

On-premise custom flow element, C#
public class SimpleFlowElement
    : FlowElementBase<IStarSignData, IElementPropertyMetaData>
{
    public override string ElementDataKey => "starsign";
    public override IEvidenceKeyFilter EvidenceKeyFilter =>
        new EvidenceKeyFilterWhitelist(
            new List<string>() { "date-of-birth" });
    protected override void ProcessInternal(IFlowData data)
    {
        StarSignData starSignData = (StarSignData)data.GetOrAdd(
            ElementDataKey, CreateElementData);
        if (data.TryGetEvidence(
            "date-of-birth", out DateTime dateOfBirth))
        {
            var monthAndDay = new DateTime(
                1, dateOfBirth.Month, dateOfBirth.Day);
            foreach (var starSign in _starSigns)
            {
                if (monthAndDay > starSign.Start &&
                    monthAndDay < starSign.End)
                {
                    starSignData.StarSign = starSign.Name;
                    break;
                }
            }
        }
        else
        {
            starSignData.StarSign = "Unknown";
        }
    }
}

The full C# on-premise example

Read the custom element example

🔗

A derived property is worked out by rules from properties that other elements have already returned, such as banding a number or combining two answers into one. Each one is a single YAML or JSON file, so adding a property takes no code.

The script below says whether a browser is a recent release. Where a property it names cannot be read, it returns no value and a message naming that property, rather than a guess.

A derived property script, YAML
Format: 1
Name: BrowserCurrency
Version: 0.1.0
Output:
  Name: BrowserCurrency
  Description: Whether the browser making the request is a recent release.
  ValueType: string
  IsList: false
  IsMandatory: true
  Category: General
  Values:
    - Name: Current
      Description: Released in the last twelve months.
    - Name: Dated
      Description: Released more than twelve months ago.
Checks:
  Known:  { Property: device.BrowserReleaseYear, Gt: 0 }
  Recent: { Property: device.BrowserReleaseAge,  Lt: 12 }
Rules:
  - When: { Failed: Checks, Eq: 0 }
    Then: Current
  - Else: Dated

Add scripts to a .NET pipeline with the FiftyOne.Pipeline.DerivedProperty package, and try one in your browser with the script tester.

Read about derived properties

🔗

You can combine our device detection and geolocation results with your internal user profile data in the same pipeline.

A single API call can return the data from the services you combine. Your application reads the combined response.

Read about Pipeline API

🔗

If you want to offer your data service through Pipeline API, talk to us about deployment and sales support.

Talk to us

Pick one to find out more

Each heading opens a short answer, with code where there is some, and a link to the documentation for more.

Read the custom element example
Why build your own?

Why build your own?

Sell your data services to others

If you want to offer your data service through Pipeline API, talk to us about deployment and sales support.

Add properties without code

Derived properties turn the properties you already have into new ones, with rules in one short file.

Streamline the data services you use

You can combine our device detection and geolocation results with your internal user profile data in the same pipeline.

A single API call can return the data from the services you combine. Your application reads the combined response.

4 ways to use Pipeline API

🔗

Use device and location data to build audiences

Group visitors by the price band and age of their devices, then add country and town from the IP address. You can target these segments without collecting behavioral data.

Who it serves. Media buyers building audiences without third-party cookies.

Talk to us about this

Pairs well with Bidstream enrichment, when you want to include these audience details in bid requests.

Integration time. One afternoon with the Pipeline API.

Ingredients, 12

Device Detection
  • PriceBand
  • ReleaseAge
  • ReleaseYear
  • HardwareRank
  • HardwareVendor
  • DeviceType
  • BrowserName
  • PlatformName
IP Intelligence
  • Country
  • Town
Bots and Humans
  • HumanConfidence

Hero ingredient. PriceBand Use the device's price band to build audience segments without collecting behavioral data.

  1. Select devices in a premium price band.
  2. Add ReleaseAge to find the flagships released this year.
  3. Add country and town for local targeting.
  4. Serve premium creative to this audience, then use 51Did to cap frequency and measure the campaign.

Chef's tip. A flagship released this year and a 2019 mid-range phone can belong in different audience segments.

🔗

Serve lighter pages to low-end devices

Cheap phones buy things too, when the page loads. Find the devices with little memory, a slow processor, or a few years of age, then send them lighter pages, smaller images, and fewer scripts. Use WebP where the device supports it.

Who it serves. Teams working on speed and conversions across global traffic.

Talk to us about this

Pairs well with Know your shopper, to learn which devices your visitors use.

Integration time. One sprint in your rendering layer.

Ingredients, 11

Device Detection
  • DeviceRAM
  • CPU
  • CPUCores
  • DeviceType
  • ReleaseAge
  • ScreenPixelsWidth
  • ScreenPixelsHeight
  • PixelRatio
  • JavascriptImageOptimiser
  • WebP
IP Intelligence
  • ConnectionType

Hero ingredient. JavascriptImageOptimiser Size images for each device automatically.

  1. Find constrained devices by memory, processor, and age
  2. Serve lighter pages, smaller images, and fewer scripts
  3. Use WebP where the device supports it
  4. Compare conversion on those devices before and after

Chef's tip. Send fewer scripts to devices with little memory or a slow processor.

🔗

Check a shared location against the IP address

Someone who's prepared to share their location, and whose location matches the IP address, gives a signal of more certainty than if there were a big divergence.

A location from the GPS that falls within the service area of the IP address is an indication that the person is actually in the country and not using a VPN.

Who it serves. Anyone who needs more certainty that a visitor is where they appear to be.

Talk to us about this

Pairs well with Suspicious traffic management, to weigh the location with the other signals.

Ingredients, 10

IP Intelligence
  • Areas
  • Country
  • LocationConfidence
  • IsVPN
  • IsProxy
  • IsHosted
Geolocation
  • JavaScript
  • Country
  • Region
  • Town

Hero ingredient. Areas The service area of the IP address, to test the shared location against.

  1. Look up the IP address for its service area and VPN flags
  2. Ask the browser for its coordinates
  3. Check whether the coordinates fall within the service area
  4. Take a big divergence as a reason for a closer look

Chef's tip. A match is a signal, not proof. Weigh it with IsVPN and Human Confidence.

🔗

Check for bot traffic in your analytics and spending

Use the crawler properties to identify visitors that give their name and purpose, then handle them according to your robots.txt. For visits that aren't named crawlers, use Human Confidence to decide what they can see and how to count them in analytics. The network properties let you review traffic from a data center, VPN, or proxy.

Who it serves. Teams checking bot traffic in analytics and ad spending.

Talk to us about this

Pairs well with Spot the fake, for a closer look at the device in a suspicious request.

Integration time. 30 minutes in any Pipeline API integration.

Ingredients, 10

Bots and Humans
  • IsCrawler
  • CrawlerName
  • CrawlerUsage
  • CrawlerProductTokens
  • IsArtificialIntelligence
  • HumanConfidence
IP Intelligence
  • IsVPN
  • IsProxy
  • IsHosted
  • HumanProbability

Hero ingredient. HumanConfidence Use the rating to decide whether a visit should see your prices.

  1. Check IsCrawler on every request
  2. Filter named crawlers by name and usage
  3. Use IsArtificialIntelligence to identify AI traffic
  4. Use Human Confidence to review visits that aren't named crawlers

Chef's tip. Crawler detection is free on every plan. Use it to handle the visitors that identify themselves before reviewing the rest.

Pick the one closest to your business challenge

The story is for you and the recipe for your engineers, with the properties they need and a link to create their Resource Key.