Member Login

The Arts360 Tool

ABA is excited to partner with HaystaqDNA and L2 to offer a first-of-its-kind tool to identify new audience members, members/subscribers and donors using predictive analytics.

 
 
 
 
 
 
 

Arts360 Resources for Current Users

Below you will find reference materials, a recording of the walkthrough, FAQs and links to helpful training videos from the tool host L2. Contact us any time if you need help with login or how to use the tool.

 
 

Frequently Asked Questions

PURPOSE OF THE ARTS360 TOOL

Q: What is the core value proposition of the Arts360 Tool?

A: The Arts360 Tool will not only enhance the consumer file around existing audience members, but the tool can also bring many new prospects into arts organizations by making use of the arts-specific predictive models.


Q: Is this tool offered outside of the US?

A: Currently, the tool itself is not offered in other countries. Arts organizations overseas can use the tool for prospecting visitors or donors in the US, however.

MATCHING  YOUR AUDIENCE  

Q: What information is needed to match lists to the L2 database? 

A: Matching requires that the upfront data you provided about an audience member is comprehensive to ensure a confident match to an individual in L2’s database — this includes looking at names, address, email, phone, gender, and date of birth.  At a minimum, first name, last name, state, and one of the following (address line, city, or birthdate) is required to do a match in the tool.

Q: Why is the number of audience members flagged in the Arts360 Tool less than the number of audience members I provided?

A: Based on the information that was given upfront in your provided audience lists, not all of your audience members are confidently matched to L2’s database. Depending on the recency of data collection and the number of duplicates, we have seen match rates from 10% to 80%.

PREDICTIVE ANALYTICS

Q: How accurate are the Haystaq arts-specific predictive models? How did you validate the data?

A: Haystaq first takes a segment of known data (i.e., audience lists initially provided to ABA) and preserves this segment as a test set to run the built models on when completed. To build the predictive scores, Haystaq runs a series of feeder models and statistical techniques — including regressions, gradient boosts, and neural networks — and finds the perfect blend of techniques. These models are then tested against the preserved segment, and in our case, Haystaq found that their models validated very strongly. 


However, the real test of the predictive models is shown when used on new data, and conveniently, because a new member matched an audience list right after the models were created, this population served as the perfect test subject. When testing to see how the dance lovers, performing arts, and donor flags showed up in the new audience list, Haystaq found that, again, the models were strongly validated. 

Q: How do the predictive flag and score interact with and relate to each other? 

A: A flag is defined as tagging all individuals above a certain score threshold. In our case, Haystaq defined the flags as the 10% - 25% of people with the highest scores (the exact threshold varied by model). We advise people to start with the flag so they get a sense of the bigger population that fits the model and then switch to the score to narrow in on a target sub-population.

DOWNLOADING DATA + PRICING

Q: Why are there small nuances in my downloaded data, i.e. I found an existing donor in a universe that supposedly excluded all of my matched audience members?

A: Because of the required data to confidently match an individual to the L2 database, not all of your audience members will have been matched into your private data (see matching FAQ section). This implies that when you’re looking at universes outside your matched audience list (i.e. prospects), you may export and capture a few current audience members who were not initially matched. 


Q: How do I download records, and what do I get? How do I get email addresses of my targeted universe?

A: View this video tutorial on how to purchase and download records. For 10¢ a record, you can receive all of L2’s consumer data around an individual (450+ pieces of information). These files come with select contact details including name, address, and phone number. The tool has pre-made file types for various purposes like mailing, or simple demographic pulls, but the tool also allows customization of fields that you wish to pull.

To get email addresses, you will select the pre-made L2 ID file type when exporting which pulls L2 IDs of your universe’s individuals — this file is how you will eventually get email addresses of your universe. Download the file (at no cost), and email the file to brynn.johnson@advisoryarts.com. Brynn will get the emails from L2 for 5¢ per record and forward them to you. Just to note, L2 has access to decades of email strings (~ 1 billion emails), most of which, however, are inactive. After running the system through a 4-stage verification process to find active emails, you’ll result in a 30-40% match of your L2 ID file to active emails.


Q: How will I get billed for the records and how will I pay?

A: Invoices will be collected at the end of the month based on records downloaded. You’ll get an invoice from ABA the same way you do for your membership fee.

L2 UPDATES + SUPPORT

Q: How frequently is the map interface and data updated?

A: Consumer data is updated on a quarterly cadence. Additionally, 3-5 million new phone numbers from 10 phone sources are pumped into the system on a monthly basis, and new emails are also collected on a 30-day cadence.

Q: What do I do if I notice a discrepancy in the interface, i.e. a typo, or a parameter count seems off? How can I pass the feedback along?

A: All concerns can be emailed to support@l2-data.com. 

Remember that not every data piece will be exactly right — people will accidentally put wrong information when filling out a loyalty card application, for example, and L2’s database is built off of hundreds of sources. 

One specific example are birthdates. L2 has a 83% match to perfect birthdates, including month, day, and year. The remainder of folks will show up in two ways: (1) 01/01/correct year; or (2) correct month/01/correct year. 

ABA HELP + ADDITIONAL SERVICES 

Q: What if I want to add a custom flag to the Arts360 Tool, i.e. flag for subscribers, donors, and STBs in my audience list?

A: With the latest update of the Arts360 tool, you can now add as many custom flags as you want. Simply match the list of individuals who meet that flag in the tool and name the flag — it will appear in your ‘private data’ section. 

Q: I accidentally downloaded records I did not mean to — what do I do?

A: Reach out to brynn.johnson@advisoryarts.com if this happens and we will work with L2 to void your account of this accidental purchase. We are here to help and understand mistakes, especially as we’re still developing this tool for your use.