Exponential Experimentation Using A Modern Data Platform

An essential part of any venture is analytics. What worked? What didn’t?

Analytics can tell you which feature to build next, which sales campaign to launch, and how much to charge for what you offer.

There can be so many places you can make changes or improvements the menu of options seems endless.

What do you do?

You split test.

For example, you decided to sell chocolate ice cream for $5 a cone in May. You sold 200 cones for $1,000.

If you want to make more money, you can make changes. You can change the flavor, add the option to have a cup of ice cream, offer discounts, increase the serving size, or dip the cone in chocolate.

Split testing comes when you change one variable in a business model to see how it impacts results.

In June, you offer fudge caramel cones, also for $5. You sell 180 cones for $900.

Now you know to stick to chocolate. Maybe next month, you add sprinkles.

Optimizing a Custom Web Application with a Modern Data Platform:

What if you add sprinkles, but to no avail? You keep trying until you finally discover the correct mix of flavor, size, and price – only to realize that you discovered the buried treasure in January.

A lot of good that does.

Can we do better than split testing? Can we change multiple factors all at once and derive a reasonable estimation of the results?

Scientists do it all the time. A/B testing in data science enables exponential experimentation.

Most businesses have so many variables they can change to tweak their sales results, they need more than a few gallons of fudge caramel to maximize your ROI.

For a greater scenario analysis, you need the correct database.

You need the proper application on top of your database to take in the data, even web scraping and data collection tools to gather it from the farthest points on the web. You need the right analytics layer underneath to digest the data your app is sending to be stored to give you a clear picture of what is most likely to work best.

One of my clients is a retailer of over-the-counter pharmaceuticals. He has a web scraping tool that crawls millions of pages for thousands of medications from dozens of his competitors.

Every day, he grabs and stores the prices, inventories, and even web traffic for product pages to see where he is competitive and where he isn’t. He stores the massive amounts of data he collects in his modern data platform to take a snapshot of the entire industry for specific medications like cold medication in November when the weather turns before we start to put on our jackets, or bandages in March for when we start exercising again after our muscles have atrophied for the winter.

He even tracks the customer reviews to see if keywords like “slow,” “expensive,” and “not available” suddenly appears in a competitor’s review.

This data is in constant communication with the analytics layer of your custom web application. Time series analysis tells him when prices change or customer interest skews. Numbers are aggregated to tally traffic numbers and even total sales for each product.

The analytics layer holds this information while the application monitors it. The moment there is a product where he offers the best price, an alert will ping him on his phone. If a competitor runs out of something, he gets the word. He gets the report if there are a series of customer reviews, all with a 0-1 star rating.

Are You Ready for the Matrix?

Modern Data Platform turns a split test scenario into a matrix of possibilities. Why alter one variable when the right custom web application can tell you what happens when you change 10?

With months and years of pricing, inventory, and customer satisfaction metrics stored in their database, they have the information to perform countless scenarios. The database can see how similar changes in specific variables in the past impacted results over the following days and months.

From that information, algorithms can be generated to answer questions about the impact of changing the price, adding inventory, or even changing suppliers.

Everything runs perfectly because all the moving parts come from the same place and run in tandem with one another. You have optimal database performance and tuning.

To get there, you need a single team to build your application and data collection tools, optimize the database, and create the analytics to track and continuously update. If the same team develops your web scraping tools, all the better.

Making your custom web application perform amazing feats like these with a modern data platform is what I do:

I am happy to talk with you about it at your convenience. For more info about the above please call us: 732-536-4765 or send us an email via our Contact Us page.

Some of the software development services we provide:

I started to develop custom software since 1985 while using dBase III from Aston Tate. From there I moved to FoxBase and to FoxPro and ended up working with Visual FoxPro until Microsoft stopped supporting that great engine. With the Visual FoxPro, I developed the VisualRep which is Report and Query Engine. We are also a dot net development company, and one of our projects is a web scrapping from different web sites. We are Alpha AnyWhere developers, and the Avis Car Rental company trusted us with their contract management software that we developed with the Alpha Five software Engine.

Comments

Search