Blogsoftware testingcustomer service

Beer, Bacon, and Big Ideas: An AdTech Team Meetup Recap

mutation testingtest coveragecookie syncingDMPsweb analyticsgeo-hashingcustomer relationsclient communicationteam buildinginternal meetuplightning talks

After an extended pause, an AdTech development team recently reconvened for the fourth edition of its informal Beer & Bacon Meetup — a recurring gathering built around lightning talks, knowledge sharing, and professional development across a range of technical and non-technical subjects.

The company's CEO opened proceedings with a lightning talk, followed by presentations from several colleagues spanning software engineering, project management, and product development.

The CEO's opening talk offered a concise and entertaining overview of cookie syncing — the mechanism by which 1st party and 3rd party cookies are used to track web traffic and share user information across Data Management Platforms (DMPs).

The utility of cookie syncing is most apparent in database-building and customer targeting, and it plays an important role in the overall efficiency of web analytics pipelines.

Dealing with Difficult Customers – Do They Really Hate Us?

Anyone who works with clients regularly has encountered the frustration of an unhappy one. Emails full of accusations and grievances are a headache no matter the industry.

A project manager on the team used their presentation to explore the underlying reasons behind these kinds of outbursts and offer practical strategies for handling them constructively.

The core insight: clients often react negatively not because of a specific failure, but because they feel a basic need is going unmet. Those needs span a wide spectrum — from financial security, which is table stakes in any business relationship, to more personal concerns like self-esteem and self-fulfillment.

Recognizing and addressing those needs doesn't just defuse tense situations — it also builds stronger client relationships and a greater foundation of trust over time.

The presenter also stressed avoiding absolute language — words like never, always, and forever — and highlighted the importance of not using the word "but" when communicating with clients, as it tends to negate whatever goodwill was just established.

Kill Them All – Mutation Testing

Testing code is a given in software development. The less obvious problem is that the tests themselves can have bugs — and standard test suites often yield green results while still failing to provide genuine coverage. That false sense of success is precisely what mutation testing is designed to expose.

A technical lead on the team walked through the approach: by first running mutants through a test suite, it's possible to verify whether those tests are actually doing their job. Common mutation operators include binary arithmetic, boolean substitution, conditional boundaries, negated conditionals, and return values.

Mutation testing does come with trade-offs — it can be costly, time-consuming, and occasionally produces false positives. But its benefits are real: it validates actual test coverage and helps eliminate unnecessary code from the codebase.

Because of the computational overhead involved, automation tooling is essentially required to make the process practical. A few options by language:

PHP

Humbug https://github.com/padraic/humbug

Python

MutPy https://github.com/bayandin/mutpy

Cosmic Ray http://cosmic-ray.readthedocs.org/

JavaScript

Grunt-mutation-testing https://github.com/jimivdw/grunt-mutation-testing

Geo-Herding – A Brief Overview of Geo-Targeting

Matching customers to relevant products and services is a foundational goal of modern advertising. Geo-targeting — using physical location as a signal when sorting and serving information — makes that matching considerably more precise.

A team of three engineers presented their work on the code underpinning geo-tracking implementations. The concept of GeoHash came up as a well-known tool in the space, though the presenters noted its limitations, particularly when it comes to calculating actual distances between two geographic points.

To address those gaps, the team shared work on refining the underlying algorithms, aiming to produce results that are both more accurate and more actionable for marketing purposes.