How to Respond to an ATO Attack
By Kevin Lee /
30 Mar 2017
So, it finally happened. One of your best customers just filed a chargeback. It wasn’t because you failed to deliver your product – you did! Actually, you fulfilled it via overnight shipping, just like your customer requested. Could it be that your buyer’s account has been taken over, and someone has placed an order on their behalf?
Although it may sting to know one of your customers’ accounts was compromised, you’re not alone. Based on Sift Science’s 2017 Fraud-Fighting Trends survey, 48% of online businesses saw an increase in account takeover (ATO) last year – and this growing threat doesn’t show any signs of slowing down.
Here are some steps you can take to deal with account takeover, both from an account-specific and holistic perspective:
What to do about that specific account
Determine whether this is truly ATO, or friendly fraud. You may assume it’s ATO, because it comes from one of your best customers – but they could just be experiencing buyers’ remorse. Although ATO is on the rise, so is friendly fraud. If it is friendly fraud, you have a better chance at winning the chargeback, especially for physical items that you delivered to the regular address on file. Therefore, make sure to save all correspondence and evidence of the transaction.
If it does turn out to be ATO, unfortunately there’s a good chance you’re going to lose this particular chargeback and be out the product that you shipped in good faith. It sucks, but treat it like an investment and an opportunity to learn, so you can prevent future instances of ATO on your platform.
Lock down the customer’s account. Once you’ve determined it’s truly ATO, lock down the account, preventing new orders from coming in until the password has been reset. Log the customer out of all open browser or app sessions – or enable view-only mode, if available.
Contact the account holder and ask them to change their password. The customer is probably as surprised about the ATO as you are, so walk them through it and use the experience as an opportunity to educate them about securing their account. If you can, have the customer enable two-factor authentication. That makes it a bit tougher for fraudsters to compromise the account again.
Put the account on a watchlist. Even after you’ve taken the steps above, it’s a good idea to watch the account and monitor their next few orders. If the customer has malware on their machine, the fraudster is going to learn any new password that the customer inputs. Suggest that your customer run an antivirus software program to get rid of malware. There are many free antivirus services out there.
Do a deep dive into the account. Investigate everything – timestamps, device, shipping address, IP, cookies, proxies, events, browser info, keylogging – to find any clues or suspicious signals that the fraudster left behind that can be used to detect other similar accounts on your platform. Take down and/or contact those customers to warn them that their account may have been compromised.
How to prevent future attacks
Dealing with the compromised account is important, but it’s just a first step. To prevent future ATO, it’s a good idea to take holistic, system-wide measures, too.
Update your models and rules with the known bad signals you’ve found. Remember that models (especially rules-based models) degrade over time, so it’s important to run periodic quality audits to make sure they’re continuing to catch bad users while not impacting too many good customers.
Educate your users. Good security practices, like using a password manager, can help protect against ATO. Publish articles in your Help Center educating your customers about account security, to help reinforce that message.
Make sure you’re tracking the right data. If your risk system is unable to monitor things like IP, cookie, device ID, session history, event velocity, and keylogging, I recommend working with your engineering team to get this prioritized. This set of data is not only useful for risk reasons, but equally as beneficial to the Marketing and Growth teams. After all, the more you understand your customers’ behavior, the better you will be at upselling them on additional products, increasing conversions … and keeping them (and your company) safe.
Re-evaluate your tools. Consider adopting third-party tools that can analyze user behavior and proactively flag anomalies.
Kevin is the Trust and Safety Architect at Sift Science. Building high-performing teams and systems to combat malicious behavior are what drive him. Prior to Sift, Kevin worked as a manager at Facebook, Square, and Google in various risk, spam, and trust and safety roles.