Finding the Sweet Spot in Your Frame Inventory: How many brands is too many?

by | Feb 17, 2022 | Data Insights and Industry Trends

Conventional wisdom

While intuition might suggest that higher frame unit sales must be supported by an increasing number of brands, conventional practice management wisdom says the opposite: less is more.  Multiple advisors on dispensary management advocate limiting brands and hence, frame vendors,,

The theory goes: by limiting the number of frame brands and vendors, ECPs can maximize discounts, payment terms, co-op funds, simplify ordering and inventory management, optimize rep service and visit time, and strengthen strategic vendor partnerships.

Can less be more? Is it possible for a practice to limit the number of vendors they must partner with to achieve high sales and capture rates?  Let’s look at the facts:

By the numbers

To see if ECPs heed the “less is more” advice and explore the relationship between the number of frames sold and ECP frame brand counts, we analyzed a sample from the GPN database containing 7+ million frames sold over the past 3 years from 2,000+ ECP locations (“doors”) across 9 US Census divisions. 

An average 147 frames were sold per door across each of the 36 months for years 2019-2021, with an average of 36 distinct brands represented. 

The correlation between frames sold and brand counts is 0.65.  This suggests an increasing relationship between frame sales and brand count.  Correlations range from -1 (a synchronized inverse relationship between variables) and +1 (a matched, same direction relationship) with 0 meaning a non-existent relationship.  

While the trend and instinct might be to expand your brand count, the data sample below suggests that a refinement of a lower brand count can yield just as high a capture rate, with the added bonus of more efficient inventory management.  

The capture rate (ratio of frames sold to eye exams) is 49% for this sample.


Using a scatter plot, we can visualize the relationship between the key indicators: brand count and frames sold. Each of the 2,000+ dots represents the coordinates of the monthly averages for an individual ECP door. The broken lines show the averages of the two measures. The intersection of the line averages divides the dots into color quadrants defined by combinations of “above” (⮝) or “below” (⮟) average for the two variables. Each quadrant represents a cluster of similar performing ECP doors.

The filter at the top, “Frames Sold per Brand per Door per Month” has 4 options as defined by 4 quartiles. The quadrant clusters, their averages and dot patterns shift with these filter choices. To un-select the filter click it again. 

  • ⮝Brands ⮟Frames: The red dots in the upper left quadrant each represent locations with the least efficient use of brand assortment.  These 15% of ECP locations have a weighted average of  2.7 frames sold per brand per month. With an average 116 frames sold per month (vs overall average 147), these doors carry 7 distinct brands more than the overall sample (43 vs 36). These locations have a big opportunity for improving assortment efficiency by reducing their number of brands. The capture rate among this cluster averages 48%.
  • ⮟Brands ⮟Frames:  The lower left quadrant’s pink dots each represent a location exhibiting an under-index performance in both frame sales and brands carried. This cluster includes 47% of the sample and is defined by a weighted average of 3.6 frames per brand. They average 26 brands sold (10 below overall average) against 85 frames sold per month (62 below sample average). Without a jump in sales, the only way to elevate their ratio of frames per brand is to be even more selective and strategic in the number of brands they carry. However, there is only so far they can go in editing their assortment, less their overall selection suffers. The capture rate averages 43% for this cluster.
  • ⮝Brands ⮝Frames: 29% of the ECP doors over index in both frame sales and brands carried, as illustrated in the upper right quadrant with grey dots.  Accompanying the highest group average of 246 frames sold per month, these locations also have the highest average of brands utilized (average 53).  They average 4.7 frames sold per brand per month and a capture rate of 57%. These locations would greatly benefit from a more disciplined approach to brand selection, which would re-position them in the good company of the next group. 
  • ⮟Brands ⮝Frames: The cluster of green dots in the lower right quadrant represents assortment efficiency with an average 8.4 frames sold per brand.  This cluster includes those ECP doors that enjoy higher than average frame sales (203 per month) from a limited number of brands (28). These exceptional practices have found the winning formula of exceeding sales averages while limiting their brand count and comprise an elite 9% of the ECP doors in the sample. Their capture rate is 56%.

Take Away

Regardless of how many frame units are sold, strategically partnering with a limited, select group of frame vendors and carefully choosing the right mix of brands maximizes an ECP’s buying power and operational efficiency. While the tendency is to increase the brand count as sales expand, 9% of doors are able to limit brands relative to sales. Their disciplined approach to inventory means efficient and simplified ordering, and likely more vendor perks.  

Quantifying and benchmarking key performance measures of frames sold, and brand counts can only be done with quality data.  Deep and robust analysis of the de-identified, anonymized GPN data warehouse enables new insights like this and identifies opportunities for industry stakeholders. 

GPN aggregates millions of transactions from thousands of eyecare providers, and focalCenter performs rigorous analysis for delivering timely and precise micro and macro dashboards with interactive market insights to the eyecare industry. Please feel free to contact us for more information on growing your eyecare business with data-driven strategies.

By Ron Krefman, OD

Finding solutions in data science.


Share This