[ad_1]
SAN JOSE, CA — Why digitally transform? Business must go through digital transformation to win share, gain efficiency, make better decisions and delight customers, said Saira Nazir, head of digital marketing at Autodesk, at MarTech Conference Thursday.
Nazir outlined three foundations of digital transformation: organizational design, data and tools and discussed how Autodesk has engaged in digital transformation to move beyond incremental KPI improvements.
Organizations must evolve. “Organizational design is probably the most overlooked” of the three said Nazir. “If you update all your tools but your teams are still relating to each other and working in the same way, you will not be able to truly take advantage of the tools.”
The approach — decentralized or centralized — doesn’t matter, the key is to focus on the evolution of your teams while going through digital transformation, she said.
More insights from the MarTech Conference
- Join us for MarTech East 2019
- Building a marketing operations team from scratch, one year in
- Why brands must take a people-first approach to martech
- It’s about teams: How Autodesk boosted conversions, retention and trimmed its sales cycle
Identify bad data. Often marketers get hung up on data gaps, but Nazir cautions that identifying bad data or metrics is critical to the process of creating effective data models and algorithmic outcomes. It’s just as important to identify and tag good metrics such as conversion rates, latency, pages to conversion, etc. as it is to identify and tag bad metrics that will pollute data models, said Nazir.
“As more companies adopt AI, bad data matters a lot,” she said. If you create a model, you need to tag good and bad data to help the machines learn and create a good algorithm. It’s not just identifying gaps in data but tagging bad groups of data.
A bad metric is “a lagging indicator instead of a leading indicator,” said Nazir. “That can prompt you go in a wrong direction.”
Scale with a CDP. To analyze and test at scale, Autodesk deployed a customer data platform (CDP).
The team saw that a large percentage of customers were spilling out of renewals and coming back in to the journey to find new pricing. “About 60% of traffic was leakage from renewals,” said Nazir. “They are already customers, and a CDP can identify them and create a much more tailored experience for those people that already have the product. We can give them tutorials to understand the tool better.” Identifying and specifically addressing the needs of these audiences increases retention.
The team used the CDP to map out the customer journey using Adobe IDs, purchase history and product attributes to understand how customers interacted with them. It then can identify appropriate audiences for tailored email campaigns based on if/and statements that direct users to specific content based on their engagements — at scale.
Chatbot to shorten sales cycle. Autodesk had 68-day purchase cycle. It wanted to shorten it and grow the percentage of people who buy online. The company looked at chatbot offerings to help surface tailored content recommendations to site visitors.
Autodesk ended up creating its own chatbot rooted in machine learning and programmed it using popular content. It ingests the referral data and answers from a couple of clarifying questions at the beginning of each session before showing content links specific to the user’s needs. Autodesk also thought carefully about the look of the chatbot’s prompts, designing it to feel familiar with the look of Instagram questions.
“We predicted customers need more information,” said Nazir. “This pushes content to them sooner.”
The initiative drove 4x more conversions and 109% more time on page, said Nazir. More impressive, Autodesk didn’t test this effort on brand keywords, but only on traffic coming in on non-brand keywords.
This story first appeared on MarTech Today. For more on marketing technology, click here.
The post It’s about teams: How Autodesk boosted conversions, retention and trimmed its sales cycle appeared first on Marketing Land.
[ad_2]
Source link
Recent Comments