The September floods in Colorado caused extensive infrastructure damage, left hundreds homeless, and destroyed thousands of automobiles. According to EQECAT, a consultancy specializing in catastrophe risk modeling, damage to the Colorado economy is expected to exceed $2 billion. Pointing to damage from the floods, a continual slow economic recovery, and shaken consumer confidence from a government shutdown, some forecasters are predicting a slow recovery for the retail sector in Colorado in the near future. While the medium and long term impacts to Colorado’s economy will take months to materialize, it is likely that local retailers will feel the pinch from these factors. The immediate questions for many of Colorado’s small business owners are:
- How soon will these economic impacts hit them and how severe will they be?
- Will my store see a decrease in number of customers (foot traffic), likelihood that a customer makes a purchase (conversion ratio), or the amount a customer spends per transaction (average transaction value)?
To answer these questions, RetailNext analyzed data from retailers in cities affected by the flooding, including Boulder, Broomfield, Aurora, and Colorado Springs, and cities indirectly impacted, including Denver, Westminster, Lone Tree, Thornton, and Littleton. Data from RetailNext customers in these areas was compared to aggregated same-retailer national data. Metrics analyzed include total traffic, net sales, net transactions, ATV (net sales/total transactions), and conversion ratios (calculated as net transactions/traffic). Our research is summarized below and is also featured on The Denver Post in an article titled, “Colorado Flood-Stricken Retailers Lost Business But Then Rebounded.”
NATIONAL & COLORADO DATA SET:
The national data set was a collection over over 450 stores comprising over 16 million shopper visits nationally with about $319 million in net sales (5.6 million transactions). The Colorado data set was comprised of stores located in the Central, Southwest, and Northwest regions of Colorado.
Metrics analyzed included:
- Total traffic, net sales,
- Net transactions,
- Average transaction values (net sales/total transactions) and
- Conversion ratios (calculated as net transactions/traffic).
- Data from several RetailNext customers in Colorado was compared to aggregated data for these stores from around the United States.
- Data included: total traffic, net sales, net transactions, and conversion ratios (calculated as Net Transactions/Traffic and average transaction value (ATV). ATV is calculated by net transactions/total traffic).
- Conversion ratio tells what percentage of the customers made a purchase (i.e. what is the likelihood that a customer who comes into the store buys something).
- Data was calculated by taking the average traffic, net sales, net transactions, ATV and conversion ratios of the previous 4 weeks prior to the flooding.
- For the flood period, RetailNext looked at 188,440 shopping trips resulting in 69,850 transactions. Here is some of the comparative data for non-flood period and national performance.
- The monthly average of a metric was calculated by using data from the previous four weeks prior to the flood. This average was compared with data for the two weeks of flooding and immediate recovery, with September 9th-16th loosely being the first week of the flood and September 16th-23rd as the beginning of recovery and this calculation was done both for the affected stores in Colorado and for the national chain over the same time period. This comparison allowed us to see:
- How much the affected store in Colorado deviated from its total traffic, net sales, net transactions, ATV, and conversion ratios baselines.
- How much an affected store deviated from national trends over the same time period.
- Allowed us to see from week 1 (the start of the flood) to week 2 (when the flood subsided) how quickly shoppers recovered and returned/continued shopping in these stores.
Pent up demand from shoppers after disaster in flooded areas
Surprisingly, in the immediate aftermath of the floods the retail sector proved unexpectedly resilient. Even in areas affected by flooding shoppers returned to stores quite soon after the disaster. Department stores received an average of 4 percentage points less traffic then the national average in the first week of the flood. Many of these stores made up for their lost customers by the next week, as midweek traffic was an average of 7 percentage points higher than the national average in the week following the flood. This increase is unusual given the floods timing in the middle of September, and illustrates that shoppers who were kept at home by pelting rains exercised their pent up demand for shopping very soon after the rains abated. Consequently, retailers who lost foot traffic as a result of the flood were able to make up some of the losses soon thereafter.
Shoppers spent more money when they went shopping
As one would expect given the extreme situation, shoppers who managed to make it out to retail stores during the week of the flood felt less inclined to make large purchases. From September 9th-16th average transaction value was 3 percentage points lower than the national average across all stores in the sample set. In the week following the flood shoppers not only came out in larger numbers, but also were more willing to spend. Average transaction value was 4 percentage points higher compared to the national average than the four week average in the week following the major flooding. A special note, we noticed that on days Monday and Tuesday, the ATV was 11 percentage points higher than the national average. The spike in ATV and traffic on the first two days of the second week when the floods subsided leads us to believe that this was due to pent up demands from shoppers eagerly waiting to return to shopping when the storm had cleared. This allowed affected retailers to recover some of their losses from the previous week, and meant that customers coming into stores were spending significantly more per transaction than normal.
After the flood shoppers were more likely to purchase
In the week after the flood, shoppers who ventured out to stores were intent on purchasing something. Conversion, defined as net transactions/total traffic, is a metric which measures how effective a store is at turning a customer that walks into its door into a sale. During the week of the flood conversion percentage was 4 percentage points below the monthly average compared to national average data. In the subsequent week, shoppers who ventured out to shop more intent on purchasing something. Conversion was 1 percentage point lower than the national average, constituting a three percent point rebound from the previous weeks low, and a number of stores had peak conversion days which were higher than any store in the country. Once they got to the store, customers were feeling more inclined to make a purchase.
Days which normally don’t receive high traffic or net sales got a boost
One unexpected result of the flooding is that pent up customer demand meant those in days following the flood which normally are slow days for retailers became relatively big shopping days. Numerous retailers had spikes in net sales on the Monday or Tuesday, traditionally slow retail days, following the flooding. These spikes ranged from 12-26 percentage points higher than the monthly average. Overall during the week of the flood, net sales compared to monthly average were 13 percentage points below the national average (change over baseline).During the 2nd week, net sales was 2 percentage points below the monthly average compared to national average data. The Monday and Tuesday were 9 percentage points above the national average. Again, this increase in the initial 2 days of the 2nd week was most likely due to pent up consumer demand for shopping.