Personalizing Customer Experience Through Dynamic Pricing in eCommerce

The vast majority of online retailers have some sort of personalization implemented in their stores. Personalization in eCommerce is usually considered to be a part of efforts to improve the overall customer experience on a website or an app. Perhaps most importantly it primarily helps with Conversion Rate Optimization (CRO) efforts.

Besides CRO, personalization can have a positive impact on customer satisfaction, engagement, and even average order values (AOV).

As we will see soon, personalization in eCommerce comes in many different forms. In this article, we will explore prices and dynamic pricing strategies as means to help with personalization in eCommerce.

Defining Important Terms

Firstly, we will go over some of the most important concepts we will use in this article, and then we will see how they act and what outcomes they lead towards when combined.

Personalization

Personalization can refer to a number of different practices. Broadly speaking, it refers to the process of tailoring each customer experience—based on data on demographics, habits, and preferences—so that each customer is presented with offerings they are most likely to engage with.

The key point that should be underlined here is data-based decision-making. Without data, there’s no personalization. There are only blind guesses (or seller’s wishful thinking) as to what offering could potentially perform well.

One simplified workflow of data-based decision-making in eCommerce would be to use Google Analytics to evaluate your product performance and use a dynamic pricing tool to set prices accordingly.

Examples range from ads retargeting, personalized newsletters (although, truth be told, the criteria for what’s called a personalized newsletter nowadays is relatively low), and relevant product suggestions to presenting customers with specific prices based on relevant data.

Pricing-related (or better yet, pricing-driven) personalization is exactly what we are going to focus on in this article. We believe that as a vendor offering price monitoring and dynamic pricing services, we can shed some light on this topic from an angle that hasn’t been explored in much detail yet.

Customer Experience

Customer experience (CX) is, to say the least, a multifaceted, complex, and broad subject. It refers to the characteristics of the relationship your buyers have with your brand throughout their whole buyer’s journey.

CX includes experience via touch points such as your actual product, your website, your salespeople, your brick-and-mortar stores, and even your ads. 

To simplify things, you can think of it as a range that on one end has frustration, and on the other satisfaction (or, in some cases even – delight). The more obstacles a customer faces during their buyer’s journey the likelier it is they will feel frustrated and have a poor customer experience. The opposite is also true, a seamless process leads to satisfaction and good customer experience. 

Here are some examples of negative and positive customer experiences:

• uninformed sales rep vs. a helpful one;
• slow and unresponsive website vs. a quick and responsive one;
• non-intuitive user flows vs. user flows created to be as efficient as possible;
• product suggestions relevant to the customer vs. irrelevant ones.

You want to have as many customers be satisfied in terms of CX because those customers will keep coming back to you and even bring in new customers. And this is, in essence, how you grow a business. 

Dynamic Pricing

We’ve now come to a topic we are most familiar with. Dynamic pricing is essentially a pricing method that aims to set prices based on real-time changes in demand. Real-time here can mean minutes, hours, days, or even weeks. 

Just as personalization, it has been made possible by a general increase in computational power, which enabled the algorithmic and automated implementation of price changes.

Furthermore, in most cases, it means automating the repricing process as much as possible. This is done by creating rules for your price changes based on various parameters you choose, some of which we will get to later on in this article. 

Another important thing to be aware of is the overall reliability of the system you’re using to deliver personalization. How consistently and precisely can you deliver on your promise to offer your customers Special offers tailor-made for them, Products they might be interested in, or a similar promise? 

The Role of Dynamic Pricing in eCommerce Personalization – Examples & Explanations

Now that we have cleared up the meaning behind these core concepts, let’s see how they work together and what kinds of outcomes can their combination bring.

To understand things better, we will take a reverse engineering approach. Let’s first take a look at some examples. Then we will break them down and see how each specific element of a personalized CX works.

Amazon

If we could have the data regarding the effectiveness of price and CX personalization through dynamic pricing, it would likely show that Amazon is the company to do it with the highest rate of success.

First of all, Amazon uses an incredibly complex algorithm when personalizing prices. Their machine-learning algorithm is based on collaborative filtering, content-based filtering, and deep learning models, among many other things. 

This is all done so that the right products would be placed in front of the right buyers at the right price. And all of this is done, as you might have guessed, in order to drive as many purchases as possible.

Besides this part of personalization, we have to talk about prices being set dynamically, as well. Depending on factors such as seasonality, competition, demand levels, and other market conditions, Amazon (and sellers on Amazon) can employ various dynamic pricing engines and algorithms to set their prices in a specific way. 

They may bundle certain products and reduce the price of the whole bundle. They may also apply a loss-leader pricing strategy where they decrease the prices of a certain product in hopes of boosting the sales of other products, which are frequently bought together. 

Uber

Uber is a peculiar case of dynamic pricing being used for the sake of something that is similar to personalization, although not quite. 

First of all, Uber’s prices are adjusted dynamically throughout the whole day, continuously. 

The most important factor for Uber’s algorithm is the time of the day, and consequently time-frames of high demand. The level of demand influences the availability of available rides, thus increasing prices.

Weather, time and distance of the route, and current traffic conditions are also taken into account. 

This all, in a way, leads to a somewhat personalized experience, when we take into account Uber’s system of personalized notifications and the fact that it takes into account factors such as user’s historical data, urgency levels, and learned preferences to tailor each user’s experience on the app.

The algorithms are constantly evolving, meaning that what’s “under the hood” of any platform’s dynamic pricing may change. 

Limited Time Offers (LTOs) with Scarcity Triggers

One clear example of LTOs combined with personalized prices at work is a limited-time offer with a special additional discount to a customer who has shown interest in a specific product.

Another example would be early access to a certain product at a discounted price for loyal customers.

Scarcity triggers such as limited quantity, exclusive access, and countdown timers can be used to increase the chances of these methods to work.

We see these methods frequently used during the holiday season or during other promotional periods such as Black Friday and/or Cyber Monday.

Do Personalized Prices Improve CX and Conversion Rates?

What must be said is that personalized prices are not the only nor the ultimate means of customer experience personalization. 

We can say that they are in certain industries a vital part of personalization, while in others they may not play as big of a role. 

We have already seen above what industries this approach is most suitable for. However, not all the industries and customers within them are so welcoming to this pricing approach. The most prominent cases where personalization is less likely to work out are:

• Essential goods and services – there’s a high chance that customers will perceive dynamic pricing in this field to be exploitative or unfair. This makes sense, because affordability and accessibility are valued much more than personalization in this case.
• Long-term contracts/subscriptions – contractual commitments usually present a legal obstacle to implementing dynamic pricing. Predictability and stability are much more appreciated and sought after.
• High-end luxury goods and services – here we are mostly talking about high-end fashion. The premium image that many brands are trying to uphold would be eroded by sharp and aggressive discounts. As opposed to personalization, exclusivity is the approach that is likelier to have success here.

Potential outcomes of price personalization

The effect of personalization through dynamic pricing will depend on the following factors:

• Do you have any other personalization efforts ongoing?
• How well do they integrate with each other?
• How well do you know each segment of your customer base?
• How will I react to my competitors’ price changes?
• How well did you perform the cost/benefit analysis of this strategy?

Depending on the answers to these questions there’s a broad spectrum of how beneficial dynamic pricing will be to you. 

The worst-case scenario would be something similar to what has happened to Wendy’s when they announced they would introduce surge pricing. They have received an enormous backlash from their customers and have withdrawn their initiative to implement this pricing approach. 

The best-case scenario is a sudden and drastic increase in customer engagement and revenue. Setting just the right prices will help you:

• Increase customer loyalty;
• Increase customer lifetime value;
• Increase share of customer; and
• Gain more market share.

Mini scenario – using dynamic pricing for personalization

Here’s a quick example of how personalization of prices through dynamic pricing would work in action.

Let’s say there are two customers browsing the website of an apparel retailer. We will call them Sarah and Alex. 

Sarah is receiving personalized suggestions based on her past behavior and preferences. This includes both the selection of the items shown to her and the prices of those items. If it exists, data from loyalty programs can also be utilized.

This means that the prices of selected items are adjusted based on, for example, profitability calculation, and a prediction of what Sarah is the most (or even least) likely to buy. This is just one example of a personalization model, we will get to them in detail later.

Alex, on the other hand, is seeing generic “suggestions”. We put suggestions in quotes because if you, as a seller, aren’t taking your customers’ preferences into account at all, are you really suggesting anything? Or is it simply a matter of chance what they’re going to see?

What’s the most likely scenario that will happen in Alex’s case? Well, based on data, Alex is most likely not going to buy anything. And that’s not even the worst thing. One session without a purchase is not something that should raise your alarms immediately.

However, over a long enough period of time, if in their browsing sessions, your visitors keep seeing only things they are not interested in, they may stop visiting your website or delete your app permanently. 

Do we even need to mention this is something you don’t want?

What you want to do in order to increase the number of purchases in your app/website through personalization is to incorporate as many relevant behavior metrics as possible in your personalization algorithm. 

Some of these metrics are (1) purchase history, (2) browsing history, (3) search queries, (4) wishlist activity, and (5) Click-through rates (CTR).

In our example, this could mean that in order to increase the chances of Alex purchasing items you could analyze and do the following:

• See what items Alex is purchasing most frequently, what items he’s purchasing together, and what items he’s purchasing rarely.
  • Once you do, create bundled offers, or discount a complementary item to the one purchased often.
• See what products Alex is frequently browsing, and see if there are any outliers, such as frequently viewed items, but with no purchases. Could it be that it’s too expensive for him?
  • Benchmark your prices against your competitors and if you are the most expensive seller, consider offering a personal discount.
• Is Alex searching for a specific category or a type of item? Do you have anything similar to offer, if not the exact thing he’s looking for? Or is your categorization not in accordance with what your visitors are expecting?
  • If you are planning to offer a missing item, make sure to inform Alex when you have it in stock. If you don’t, consider creating a suggestion for a different item that would be based on what you can offer and what has the highest potential of being interesting to Alex.
• Are there any items in Alex’s wishlist that have been there for a long time? Could you consider presenting them when they add an item to the cart, or when they are about to checkout? 
  • Experiment with various UX solutions (depending on your budget) and see what brings in the best results.

Additional Tips & Tricks for Better Personalization

Now that we have covered the major part of the discussion around dynamic pricing and personalization, let’s look at some fundamental prerequisites and even some more nuanced techniques related to this approach.

Set goals (KPIs) and methods to track & evaluate them

The Key Performance Indicators (KPIs) should be set in such a way that the objectives of your personalization efforts are clear and measurable.

One popular approach is to set the SMART goals, a concept first proposed by George T. Doran in November 1981. SMART stands for:

• S – specific,
• M – measurable,
• A – achievable,
• R – relevant,
• T – time-bound.

One immediate candidate for a such KPI is—you’ve guessed it—revenue. Along with it usually go conversion rates, average order values, and return on investment. These are usually considered core KPIs of almost any marketing or sales effort. 

However, you shouldn’t stop there. 

It’s almost as equally important to track other metrics such as customer retention rate, net promoter score, and overall satisfaction ratings, and listen to other types of feedback coming from your customers.

Never underestimate honest customer feedback as a source of new ideas for your business.

Continuously test different dynamic pricing algorithms

There’s not one penultimate dynamic pricing algorithm that will be the end of your work on your prices. 

As market conditions and your goals change, so will (or, at least, should) your goals, strategies, tactics, and ultimately your pricing rules.

At a certain point, you may wish to see if you can drive more purchases by stimulating a less active segment of your customer base with aggressive discounts.

At other times, you may not want to enter a pricing war with a competitor and you may stop the algorithm from further decreasing your prices. 

Another possible scenario is that you may want to upscale your brand and start applying premium pricing. This is something Coach has done over the years. 

Communicate appropriately with your customers and listen to them

Communication is a key part of almost any personalization effort. 

It directly plays into your customers’ need to be listened to. By showing that your personalized prices and offerings are intentional, you improve your brand’s perceived value in the eyes of your customers.

How to do it right depends on each specific brand. How are your brand’s voice and tone defined? How does your visual identity play into your communication efforts? Are your customers coming from different cultural backgrounds and require tailored communication approaches?

The answers to these questions will vary greatly depending on where you are located, what customers you are serving, and what industry you are in.

Conclusion

As we have seen, dynamic pricing can be a powerful tool when personalizing CX through prices in eCommerce.

If you are already familiar with dynamic pricing software, brainstorm some ideas as to how it can be used with other personalization algorithms you are already using.

This article can be a starting point to give you a conceptual starting point and some ideas as to how you may apply this approach in practice. 

We have already mentioned, but careful experiments are key. 

See what your customers react positively to and also what they stay away from. Keep testing one parameter after another and see what is the most optimal approach.

The complicated thing here is that the answer changes based on numerous factors. What works in one industry may not work in another. What works during a certain season, may prove to be disastrous during another. 

The key takeaway is to come up with ideas, test them, and react to changes quickly, which is something that automated and algorithmic tools will help you with.