Frequently Asked Questions

• Is this right for my business? (our qualification criteria)

Longbow works best for companies with more than 500 customers, with two or more products and with more than one transaction per customer for many customers. Some existing users have millions of customers and over 100,000 sku’s.

• How does the program work?

First you sign up and give Loyalty Builders a purchase order covering the amount of information you expect to download. Next you use the Data Upload tab to extract the relevant transaction data from your files and send that data to Loyalty Builders for analysis. When the analysis is done, you get an email notification. Then you can go to the Reports tab to see a high level view of your business.

You are now ready to use the Customers tab to create lists of prospects and view the results of campaigns using previously created lists. After you use the selector to set criteria for a list select, you download the list of names and selected data for use in a campaign.

The next time the data is analyzed, you can see activity summaries for customers whose names you have downloaded.

See the Quick Start Guide for more details.

• What does it cost?

There are two pricing plans – subscription and pay as you go. In the subscription plan you pay a fixed, agreed-upon amount per analysis. In the pay as you go plan, you are charged according to the number of records you download, at $100 per $1000 records.

• How secure is my customer data?

When your data is uploaded to Loyalty Builders, the transmission is encrypted.

When your data is stored on our servers, it is also encrypted.

If you send only a customer ID for each account, no one would have any idea who the customer is. If you send both ID and name and address information, your output is more understandable, but it is then important to protect the login information for your account.

• How long does it take before we can see the results of the analysis?

For the first data set, allow about one to two weeks to iron out any questions about product names, sku’s or other data fields peculiar to your company such as your designations for sales channels. Thereafter, expect to see analysis results within a day or two of data upload.

• Why is there a campaign calendar?

Purchase probabilities are perishable and change over time. By specifying when you intend to campaign against the prospect lists you download, you get names of customers likely to buy in the specified period.

• What kind of data do we need to send to Loyalty Builders?
   How do we send it?

You upload transaction data: customer ID, transaction date, transaction amount and descriptors that identify the product sold. The Data Upload tab is a program that extracts the necessary records from your ERP data files and automatically transmits the records to Loyalty Builders via File Transfer Protocol (FTP) over the Internet.

• How do we measure the success of campaigns we run based on these predictions?

You should look at the uplift in response rate and revenue compared to your business-as-usual method. Longbow helps you do this by reporting campaign results as part of the analysis associated with the first data upload after the end of the campaign.

Why are there so many choices for the probability threshold? Won’t we get the best results by setting the threshold high and just marketing to customers with high probabilities of purchase?

The higher you set the probability threshold, the higher will be the response rate percentage. However lowering the threshold will typically yield more buyers and more total revenue along with a lower response rate. Using the campaign selector as a ‘what if’ tool will let you find the optimum threshold.

• How often should we submit our data for an analysis?

The frequency of data submission depends entirely on the frequency you need to have the predictive analytics refreshed. For most direct marketing campaigns, quarterly will be sufficient. However there are many clients who require a monthly refresh of the data.

• How do I locate the necessary transaction data in my systems?

Four kinds of data are needed: a customer ID, transaction date, transaction amount and one or more product identifiers like SKU and product group. The records with this data are typically in sales order processing tables.

• How can I tell if an FTP link is established? Will I know when my data is received?

You will get on-screen messages verifying the link and notifying you when the data is received at the Loyalty Builders server assigned to your account.

• How do I interpret a Velocity Map?

Each velocity map (like the one in the Reports tab) shows a series of colored dots. Each dot is an identifiable customer characterized by three numbers. One number is the position of the dot on the horizontal x-axis, for example the change in revenue from one period to the next. The second number is the position of the dot on the vertical y-axis, for example the change in Loyalty Score from one period to the next. Finally, the color of the dot corresponds to the Loyalty Group of the customer. Lgroup 6 is composed of the best customers.

• What is a Back Test? How do I interpret the Back Test chart and table in Reports?

A Back Test is a measure of the validity of the predictive analytics. Predictions are typically made for the most recent period, the period that concludes with the latest analysis. The predictions are made by figuratively stepping back in time to the beginning of the most recent period and then predicting what will happen in the most recent period. The predictions are then compared to the actual results from ‘business-as-usual’ marketing.