4 Marketing Personalization Mistakes to Avoid

girl-covering-face
We’ve posted quite a few articles on marketing personalization best practices and ways to increase value with personalization, but what we haven’t touched on are things to avoid when it comes to building momentum with marketing personalization and automation.

Here are four marketing personalization mistakes you absolutely have to avoid like the plague if you want a smooth ride(relatively) on your path to personalization.

1. Infringing on customer privacy and not protecting customer data

Don’t be manipulative when it comes to gather information from customers

All it takes is one screw up for a huge PR disaster and plenty of lost potential and current customers. Just don’t do it.

This applies to email opt ins on retailer websites, to mobile app permissions, to social log ins on websites. Be clear, respect your customers by letting them know exactly what you will be using information for, and you’ll earn their respect.

Since one of the first steps of true hyper-personalization is building an integrated data management system that can bring in multiple external and internal data sets, the inherent risk is quite clear. With all your data in one location, there must be significant care in protecting the customer gold harvested because one data breach can mean multiple streams of data are vulnerable.

Be honest with your customers about what you are taking from them, and once you have their trust, protect what you have. It’s that simple.

2. Relying on one set of data

To build a 360 degree view of your customers, you need to draw insights from various data sources. While one data source may constitute a large majority of your data analysis into your personalization platform, the more diversified your data collection points are, the more accurate your predictive analytics will be.

For instance, a big box retailer with may point to POS data and their eCommerce data as their main data feeds into a marketing personalization tool, but forgetting to integrate social media data for crucial life event data would be simply be a waste. There will be sources of data that will be more relevant than others, but finding out where to piece in and weigh each data channel is too important to ignore.

3. Neglecting testing

Testing is a pain. Multivariate testing can get very messy with hyper-segmentation, but always remember to test while executing. The closer you get to hyper-personalization, the more marketers will be tempted to skip various parts of the testing process.

Don’t fall into that trap. Just because the testing process will become more complicated doesn’t mean you should take your foot off the testing pedal. It will become even more important to your personalization journey that all your data sources, creative pieces, and messages are carefully tested to optimize your personalization efforts. Remember that a marketing personalization tool is exactly that…a tool that needs constant recalibration to make the high, consistent returns that you expect.

4. Thinking you’ve reached true hyper-personalization

Thinking that hyper-personalization is a place where you will someday reach and lay claim to is unreasonable and dangerous to long-term marketing personalization efforts.

Algorithms can be update and tweaked, new sources of data can be added, execution points can be refined and tested.

Knowing your customer 100% and predicting their needs exactly won’t happen without having Jedi mind reading powers, but you can always keep moving in the right direction.

Nobody said personalization was easy, which is why very few have figured out the right path towards marketing personalization. With these tips in mind, you’ll save yourself a lot of time and money while consistently moving and accelerating in the right direction.

Stop Dirty Data: 5 Steps to Data Cleaning

One of the biggest issues we see our clients have is dirty data; that is, inaccurate, erroneous, or otherwise incomplete data. While this may seem like a tiny blip in the realm of big data and analytics, it is quite possibly the biggest barrier in enhancing a company’s overall data insight capabilities. Here are 5 steps to data cleaning and ensuring data integrity.

1.) Know Your Data

Before you even attempt to start cleaning up your data, you need to ask yourself: What is the data used for currently and how it will be used in the future? Without a direction of how the data will be used, it will be difficult to clean it up because you won’t know what fields to eliminate (if any) or understand what glaring gaps lie within the data. We were given a task from by a client to clean up their data and bring it all in one place. When we presented the pristine, cleaned up, data it to them, we asked them what it would be used for and their response was they had no idea!

Bottom line: if you don’t know what your data will be used for, you will have a much harder time knowing where to even start the clean-up process.

2.) Sanitize Your Inputs

Let’s talk about how your data ends up in a database. Jane registers for a website and enters her name, e-mail address, and zip code. Her e-mail address is stored within the database with an extra space due to the input of the form, so instead of “janedoe@mail.com”, it reads as “janedoe@mail.com “. While this may seem insignificant, having that extra space will make the email field unreliable and could cause problems with how the data is read in other programs. Cleaning up your currently stored data is all well and good, but if you haven’t purified your inputs, there will be a continuous loop of bad data.

3.) Identify a Unique Field

This may seem like a no brainer, but without a unique field, harvesting your dataset will be much more difficult. Databases such as SQL and MongoDB insist on storing a unique field, labeled Primary Key or Object ID respectively. This unique field is important in identifying links between two tables, so you are able to append data from one table source to another.

For example, let’s say you have two separate tables: your customer list and your e-mail campaign data. Each customer has a reference number which is located in both tables. In order to link the customer data to the e-mail campaign data, having the unique reference number is vital in connecting the two tables.

4.) Validate Your Fields

If you pay attention to only one step in this article, this should be it. Validating your fields is vital in cleaning up your data.

Remove Spaces

Having spaces in your data could directly affect the readability if another software or API needs to read the information. Be proactive in removing any unnecessary spaces in your fields. If there is truly a space in the cell, the =TRIM function in Excel eliminates the unnecessary spaces.

Remove Illegal Characters

Sometimes illegal characters make their way into your otherwise clean data. We all know naming file names on your computer with characters such as #$%^& will elicit an error response, and the same goes with cell data. For example, let’s say you have a “date” column, and for whatever reason the “#” symbol is being inserted in every row. The date column will not read as a date column but as a text or general column.

Do Not Store the Date Field as a Text String

The date column is crucial in determining when an event took place. Let’s say you want to view the revenue for the past year that were attributed to your e-mail campaigns. If the e-mail sent date field is reading as text and not an actual date, you won’t be able to filter the data to gather your revenue.

Store E-mail Address as Lowercase

For consistency purposes, it’s important to store your customers’ e-mail addresses as lowercase. If you store a customer’s e-mail address as uppercase and he/she then decides she wants to unsubscribe from communications, the e-mail address may not get unsubscribed because of case mismatch.

5.) Keep Tomorrow in Mind

Looking ahead when it comes to data organization and cleaning will save you valuable chunks of time in the future. Not all data fields may be useful immediately, so hiding and safekeeping seemingly useless information instead of deleting fields will help save time and money when business needs change. Process guidelines for data structure guidelines may be time consuming to make, but it will mean a world of difference when explaining data entry and manipulation to new hires and other team members.

Consistency is the key to a cleaner data future.