All Collections
Sniffie: Customer's Help Collections
Pricing
Price optimization - AI
Insights FAQ
Insights FAQ

This article addresses some of the most common questions related to the insights in Sniffie enterprise app

Niko Naakka avatar
Written by Niko Naakka
Updated over a week ago

Table of contents:

  1. What are insights?

  2. Where can I access insights?

  3. List of all insights not requiring any AI module

  4. List of insights requiring one or more AI modules

    1. Elasticity insights

    2. Stock insights

    3. Discount insights

    4. Rank insights

    5. Price increase insights

    6. Percentile insights

    7. Percentage of maximum insights

    8. Insights under normal filters

    9. Other insights

What are insights?

Insights is a module in Sniffie which allows you to easily find out which actions would be most beneficial to perform on any of your products. Please note that insights are dependent on what customer data is available in Sniffie.

Where can I access insights?

You can access insights from your Product Catalog view by clicking on the filters button.

If you have the insights module enabled, you will see a separate section for insights below your filters when opening the side menu. Insights behave similarly to filters in product catalog.

In dashboards, you can use insights to filter widgets and in pricing you can use insights with automatically updating pricing strategies. In price matrix, you can also use insights with the view.

List of insights which do not require an AI module

  1. Product has AI suggestions: Product has an AI suggestion regarding what to do with the product.

  2. Not meeting margin target: Product's current price is below the set margin target.

  3. Product loses money: Product's current price is less than its costs (including tax, if applicable).

  4. Missing required fields for AI: Product has certain fields that are missing or invalid (usually cost is undefined, or price = 0).

  5. Pricing strategy execution fails: There is an unrecoverable failure in pricing strategy execution. Click on the product catalog row, and then on the product page click the notice and error log to see what the error is.

  6. Not sold during last 1 year: All products that have not had a single transaction within last 1 year

  7. Only single price recorded: All products that have had the same price all the time in the transaction / price history.

  8. Last price change days ago: The number of days since the last price change. This value does not come from the transaction history, but via the price history. If no change, value is not recorded.

  9. Last price increase days ago: The number of days since the last price increase. This value does not come from the transaction history, but via the price history. If no change, value is empty and not shown.

  10. Last price decrease days ago: The number of days since the last price decrease. This value does not come from the transaction history, but via the price history. If no change, value is empty and not shown.

List of insights requiring one or more AI modules:

Elasticity insights:

  1. Product is elastic: Product seems to behave elastically (i.e. volume chances with price).

  2. Product is inelastic: Product seems to behave inelastically (i.e. volume does not change with price).

  3. Volume increases with price: Volume seems to increase with price. This can happen if the product has behaved inelastically before and is not transitioning to elastic. Check the volume vs. price graph for the product in question.

  4. Percentage to profit optimum: How far in percent is the product from currently estimated profit optimum range (within 3% of profit optimum).

Stock insights:

  1. Run rate in days: Indicates how long in days the current stock lasts. Calculation is done utilizing that product's seasonality (with all applicable seasonality components: weekday, week number, monthly, quarterly, yearly and linear), current AI result as well as current available stock for the product. We extrapolate from the current stock with the AI result (finding out the volume for that price) and the projected seasonality modifier for all upcoming days within the next 3 years and find out in how many days the stock runs out (i.e. stock run rate). If no seasonal components are available, it is assumed that product is non-seasonal.

  2. Days since last purchased: Cutoff for days since last purchase has been recorded in our system.

  3. Average days per one sale: How many days, on average, it takes for one sale to occur. For this insight to be available, we need at least 6 months of product sales history (including 0 days) since otherwise this can be inaccurate.

Discount insights:

  1. Benefits from a small price drop: Product seems to benefit from a minor price (<10%) drop. These may be good candidates for promotional coupons.

  2. Benefits from a medium price drop: Product seems to benefit from a medium (10-20%) price drop. These may be good candidates for supporting products in discount campaigns.

  3. Benefits from a large price drop: Product seems to benefit from a large (20-30%) price drop. Consider putting them on a discount campaign.

  4. Benefits from a very large price drop: Product seems to benefit from a price drop > 30%. Consider putting them on a discount campaign.

  5. Campaign days used this year (running 1 year): On how many distinct days the product has been marked as being in campaign in the price history for the last 365 days. Only observes days from which the integration has been live and cannot extrapolate on any days before that.

  6. Campaign days used since January 1st: On how many distinct days the product has been marked as being in campaign in the price history since January 1st of the current year. Only observes days from which the integration has been live and cannot extrapolate on any days before that.

  7. Number of days in normal price this year (running 1 year): On how many distinct days the product has been priced with normal price (i.e. non-discounted) in the price history for the last 365 days. Only observes days from which the integration has been live and cannot extrapolate on any days before that.

  8. Number of days in normal price since January 1st: On how many distinct days the product has been priced with normal price (i.e. non-discounted) in the price history since January 1st of the current year. Only observes days from which the integration has been live and cannot extrapolate on any days before that.

Rank insights:

  1. Sales quantity rank: When all products are put into line from highest quantity seller to lowest quantity seller, what is the absolute position of the product. Lower number is better.

  2. Revenue rank: When all products are put into line from highest revenue seller to lowest revenue seller, what is the absolute position of the product. Lower number is better.

  3. Profit rank: When all products are put into line from highest profit seller to lowest profit seller, what is the absolute position of the product. Lower number is better.

  4. % of total quantity: What is the percentage of total quantity that the product has sold out of all sold quantities the store has sold.

  5. % of total revenue: What is the percentage of total revenue that the product has sold out of total revenue the store has seen.

  6. % of total profit: What is the percentage of total profit that the product has sold out of total profit the store has seen.

  7. Daily revenue (avg): Average daily revenue the product brings per day. Calculated based on any days it was available for purchase.

  8. Daily quantity (avg): Average daily quantity the product brings per day. Calculated based on any days it was available for purchase.

  9. Daily profit (avg): Average daily profit the product brings per day. Calculated based on any days it was available for purchase.

Price increase insights:

  1. Benefits from a small price increase: Product seems to benefit from a minor price (<10%) increase.

  2. Benefits from a medium price increase: Product seems to benefit from a medium (10-20%) price increase.

  3. Benefits from a large price increase: Product seems to benefit from a large (20-30%) price increase.

  4. Benefits from a very large price increase: Product seems to benefit from a price increase > 30%.

Percentile insights:

  1. Percentile of revenue: Gives out the percentile of revenue, where 0% is the top seller and all other products are ranked behind it in order.

  2. Percentile of profit: Gives out the percentile of profit, where 0% is the top seller and all other products are ranked behind it in order.

  3. Percentile of quantity: Gives out the percentile of quantity, where 0% is the top seller and all other products are ranked behind it in order.

Percentage of maximum insights:

  1. Percentage of max revenue: Gives out the percentage of revenue when compared to the product with maximum revenue. 90% means that you receive 90% of the revenue of the top performer product.

  2. Percentage of max profit: Gives out the percentage of profit when compared to the product with maximum profit. 90% means that you receive 90% of the profit of the top performer product.

  3. Percentage of max quantity: Gives out the percentage of quantity when compared to the product with maximum quantity. 90% means that you receive 90% of the quantity of the top performer product.

Insights that you can find under filters instead of insights

  1. ABC analysis (revenue): Standard ABC analysis done on revenue. A is better than B which is better than C. We also inform on the subclass 1-3, where 1 is the best and 3 is the worst. Thus A1 are the best performing items, and C3 is the worst.

  2. ABC analysis (profit): Standard ABC analysis done on profit. A is better than B which is better than C. We also inform on the subclass 1-3, where 1 is the best and 3 is the worst. Thus A1 are the best performing items, and C3 is the worst.

  3. ABC analysis: (quantity): Standard ABC analysis done on quantity. A is better than B which is better than C. We also inform on the subclass 1-3, where 1 is the best and 3 is the worst. Thus A1 are the best performing items, and C3 is the worst.
    ​
    All ABC analysis classifications are made with the following formula:
    A1 < 0.067
    A2 < 0.134
    A3 < 0.2
    B1 < 0.3
    B2 < 0.4
    B3 < 0.5
    C1 < 0.67
    C2 < 0.84
    C3 rest
    So for example if product is in the top 0.05 (i.e. 5%), it is A1, if it is in the top 0.6 (i.e. 60%), it is in C1.
    ​
    ABC analyses are based on latest year's sales (rolling 1 year backwards, not the last calendar year). The classification is only done if the product is currently visible in the app.
    ​

  4. Peak month: The month number (1-12) when the product is expected to sell the most if the product is seasonal. This has value n/a if product is not seasonal or there is not enough accuracy for the seasonality.

  5. Peak month start date: If product has peak month (see above), this is the first day of the next peak month it has, and thus available for date filtering.

Other insights:

  1. Seasonality info available: Product has enough data for full seasonality that can be used in forecasting modules in its greatest detail.

  2. Coming into season soon: A seasonal product is currently out of season it's season will start in the coming month(s).

  3. Going out of season soon: A seasonal product is currently in season but it's season is ending in the coming month(s).

  4. In season now: A seasonal product is currently in its season (i.e. sales expected to be high for its reference level).

  5. Out of season now: A seasonal product is currently out of season (i.e. sales expected to be low for its reference level).

  6. Does not appear seasonal: Product does not appear seasonal or the accuracy of the seasonality is too low.

Did this answer your question?