Hi {{firstname|everyone}}, 

There’s a part of auditing that everyone just accepts as a cost of doing business:

Pick a few hundred transactions. Hope they represent the full picture. Burn through hours testing line items that might not even matter. 

Still most firms do it.  

But when you actually pause to ask why, you realise we’re throwing people, hours, and money at the wrong end of the problem. 

That’s why stories like Cherry Bekaert’s are so important. Not because they’re using some shiny new tool. But because they’ve shown that the status quo is costing us more than we think. 

 

Sampling in Audit Was Built for a Time Before Tech 

Traditional sampling was built for a time when reviewing every transaction wasn’t feasible. So, we relied on statistical methods to offer reasonable assurance within tight time and resource limits. 

Today, AI tools can scan entire ledgers in seconds 

Risk profiles can be built based on historical patterns. And anomalies can be flagged across 100% of data. 

But the truth is, only 15% of mid-sized audit firms currently use AI to support audit risk assessments. 

Most audit firms still follow the old routine: 

  • Pick a sample based on size thresholds. 

  • Test every item, regardless of materiality or risk. 

  • Spend hours on transactions that are likely low-risk and irrelevant. 

The result? Audits that are slower, more expensive, and ironically, less risk sensitive. 

That’s what Cherry Bekaert tackled head-on. By using AI to assess risk across the entire dataset first, they reduced their sample from 384 to 252 transactions, not just for efficiency, but to zero in on where issues were most likely to arise. 

 

Audit Hours Are a Black Hole for Time and Margins 

Audit work eats time. Time eats margin. And the only way most firms have found to manage it is by throwing junior staff at it or raising fees, neither of which clients appreciate. 

EY’s Global Assurance Innovation Report 2023 revealed that firms that implemented AI-based audit tools reported a 30–50% reduction in time spent on routine testing tasks.

Verifying 300+ random transactions, documenting support, checking for compliance… It’s repetitive, rule-based, but exactly the kind of work AI is good at. 

That’s how Cherry Bekaert reduced their sample size.  

They saved about $4,400 per engagement. That’s not because they cut scope or compromised quality. It’s because they stopped wasting hours on low-risk, low-yield work. 

Most firms say they want to become more advisory. But unless they reclaim their time from these operational sinkholes, they’ll never get there. 

 

Clients Want Confidence — Not 350 Random Line Items 

The very thing we do to give clients confidence — audit sampling — can make them doubt us. 

According to PwC Global CFO Survey, 2023, 78% of CFOs now expect their auditors to use advanced analytics or AI in engagements. 

We’re in a new era of client expectations. AI-backed risk assessment helps shift that conversation 

Cherry Bekaert used this approach to make their audits more credible, not just more efficient. 

And that’s where the real competitive edge lies. 

 

What We’re Doing at Samera 

At Samera, we’re seeing the same shift across our clients and within our own firm. 

We work with accountancy firms that are stretched thin, buried in manual work, and under pressure to deliver more value without more headcount.  

What’s more, you can’t solve modern problems with legacy systems. 

That’s why we’re building our own platform, Samera AI, which employs AI-assisted workflows that help: 

  • Automate sampling logic in audits 

  • Risk-score transactions before human review 

  • Free up staff time for higher-value judgment work 

Want to know more about the synergies between AI and accounting? Book a call today: 

Cheers, 

Arun 

Keep Reading

No posts found