Skip to main content
Analytics metrics overview

Sources, calculations and additional information about all metrics available in the Connecterra Platform

Updated over a year ago

Feed metrics

Source: All metrics are sourced from the feed software. The number of animals per pen is sourced from the DHMS (we assume that this number is more accurate in the DHMS than in the feed software).

Filters: You can view averages per pen, status, production group and custom production group. Learn more about these segments here

Available metrics:

  • Dry matter intake – this is the daily delivered feed minus the refusals (only if the refused feed is weighed and entered into the feed software).

  • Feed cost per head - average cost of delivered feed minus the refusals (available if the costs per ingredient are added in the feed software)

  • Refusals - average daily weigh-back (sum of refused feed) divided by the total weight fed. Note that this data is only correct when all refused feed is weighed and entered into the feed software.

  • Feed efficiency from milk yield - daily milk yield (from milking parlor) divided by the daily dry matter intake (minus the refusals, from feed software). *Only applies to milking cows

  • Feed efficiency from ECM – daily ECM yield per cow divided by the daily dry matter intake per cow (minus the refusals, from feed software). ECM is calculated from the daily milk yield (milking parlor) and fat and protein % (milk processor). *Only applies to milking cows

  • Feeding time & first feeding time - the average feed delivery/first feed delivery time. The average feeding time accounts for multiple feedings per day (for example: if a pen is fed at 4.00 AM and at 2:00 PM, the average feeding time will be 8.00 AM. The first feeding time will be 4.00 AM). For production groups, the first feeding time will be the earliest among all the pens in that group that day (for example: the High Production Cows pens are 2, 3, 5, 6 and they were fed respectively at 4.13AM, 5.16AM, 4.45AM and 4.01AM, the value for that day will be 4.01 AM).

  • Loading error – average loading error of all selected pens, per day. If there are no pens selected, it's the average error of all pens.


Behavior metrics

Source: The platform integrates behavior data from third-party activity sensors such as: Datamars Livestock, Nedap, Lely, Cowmanager, Heatime (Sensehub) and more.

Filters: You can view average values per pen, DIM group, reproduction stage, lactation number, production group and custom production group. Learn more about these segments here

Behavior patterns

Aside from the charts showing average time per day for each behavior, you can visualize daily and weekly patterns in cow behavior with heatmaps. This allows you to spot issues and opportunities to improve consistency.

Reading the heatmaps: Each tile represents an hour and the color indicates what % of the hour the cows spent on a given behavior - the darker the color, the higher the percentage.


Production metrics

Source: The platform integrates production data from different sources:

  • Herd milk test – this data is sourced from the DHMS. The frequency of the herd test varies per farm. Whether this data is available depends on the DHMS integration.

  • Milking system – this data is measured daily by the milking system and sourced from the DHMS. Availability of daily milk yield data depends on whether the farm has milk meters integrated with the DHMS and (for ECM) whether shipped milk data is integrated into the platform.

  • Shipped milk - this data is sourced from the milk processor. We retrieve this data for most processors daily, in most cases, for the last 14 days. We use deduplication to ensure we don't show any duplicate data. We use the ticket number to check if a record already exists, overwrite or add data to this record.

*If your milk records don't have ticket numbers, we will use the date + milk weight as identification. This can cause issues: if the milk yield changes, we could add a duplicate because we have the corrected and the old value.
Did you spot an issue? Please report the issue via the feedback button and mention what date the issue occurred.

Filters:

Herd milk test data is available per pen, DIM group and lactation number.

Milking system data is available per DIM group, lactation number, and for some metrics (milk yield and ECM) per pen and custom production group.

Processor data is only available per farm. Learn more about these segments here

Available metrics:

  • Milk yield per cow - is the average daily production per lactating cow. This metric can be calculated from different sources:

  1. Milk yield (from milking system) - the average yield per cow from daily milkings

  2. Milk yield (from herd test data) – the average yield per cow on herd test day

  3. Shipped milk yield (from processor data) - is calculated by dividing the milk delivered on a given day by the number of lactating cows on that day. These are the weighted averages of all milk deliveries per day. This metric is adjusted using a data science model.

  • Shipped milk (total) - the sum of all milk deliveries per day, from the milk processor. If there is a day with no milk deliveries, there will be no data point for this day.

  • Energy Corrected Milk:

  1. ECM (from herd test data) = 0.327*milk yield (herd test)+12.95*fat yield (herd test)+7.65*protein yield (herd test)

  2. ECM (from daily milking and shipped milk) = 0.327*milk yield (milking system)+12.95*fat yield (shipped milk)+7.65*protein yield (shipped milk)

  • Fat percentage – weighted average fat %

  1. De novo fatty acids are synthesized in the mammary gland and their increased concentrations are correlated with overall milk fat content. The generation of these fatty acids requires a primer, one of which is beta-hydroxybutyrate (BHB), an indicator of ketosis. This is why cows with ketosis often have higher milk fat levels.

  2. Mixed fatty acids can originate from both preformed and de novo sources, but they cannot be differentiated because they are long-chain fatty acids.

  3. Preformed fatty acids are transferred directly to the mammary gland from blood plasma, with more than 75% coming from the diet. Effective forage management can help promote higher concentrations of preformed fatty acids. The remaining portion is mobilized from body lipids, and this proportion can shift temporarily after calving.

  • Protein percentage – weighted average protein %

  • Somatic cell count – number of white blood cells in milk. This indicates the udder health and consistency of prep during milking. Note that all SCC values are x1000 cells/mL.

  • Solids – calculated by multiplying the fat and protein percentage by the yield

  • Conductivity – can reflect a higher salt percentage in the milk, which indicates a higher risk of mastitis.

  • Fat-to-protein ratio – the ratio of fat and protein percentage in the milk

  1. A ratio of >1.45 indicates a higher risk on ketosis (in certain (DIM)-groups)

  2. A ratio of <1 indicates a higher risk on acidosis (in certain (DIM)-groups)

  • Milk Urea Nitrogen (MUN) - indicates the nitrogen efficiency in the ration. Recommended values differ according to different researchers and lie somewhere between 8-14 mg/dl. The main factors that affect this are the ration and the way of feeding, but also the time between feeding and milking, the cow breed and the number of milkings per day. Low values (<8-10 mg/dl) indicate a possible dietary protein deficiency; source: Interpretation of Milk Urea Nitrogen (MUN) Values (psu.edu).

What are the “adjusted” metrics?

We use a smart data model to adjust the data provided by the milk processor, filling in data points on days with no milk pickups to make charts easier to read.

The % fat (adjusted), % protein (adjusted) and SCC (adjusted) charts show data completed using the interpolation method. Put simply, this means we use known data values to estimate unknown values.

The Shipped milk and ECM (adjusted) chart shows more reliable values, as the new model accounts for the days when no milk was delivered and the variation in the number of milk trucks per day.


Herd composition metrics

Source: The metrics are sourced from the dairy herd management system and include all reported animals.

Filters: You can view averages per pen, DIM group, production stage and lactation number. Learn more about these segments here

These metrics include the number of animals on the farm, their distribution across different groups, as well as stocking density numbers.

For stocking density metrics, the farm owner/herd manager needs to report pen capacities (number of stalls and headlocks per pen) to the Connecterra team. *If there are no headlocks, we use the length of the feed bunk and allocate 2 Feet per cow.

If we don't have capacity data for a pen, it will be excluded from the calculation. If the capacity of a pen has changed, please get in touch with us to update the calculation.


Weather metrics

These metrics are measured outdoors, based on the farm's GPS coordinates and the nearest weather station. They are sourced from Visual Crossing Weather API.

Temperature Humidity Index is an indicator related to heat stress for cattle. It is calculated based on the temperature and relative humidity. With a THI above 72 cows will likely experience heat stress that will impact reproduction. Above 78, milk production can be seriously affected (Vitali, 2009).


Finance metrics

Milk price – we currently use Class III USDA milk price

Income over Feed Cost (IOFC) gives you a clear overview of your farm’s profitability, allowing you to evaluate and optimize management decisions.

IOFC is available for farms that have data on per-head milk pickups and feed cost. It is shown in dollars per cow per day and is calculated using the formula below:

IOFC = milk pickup weight (per head) * milk class price - feed cost per head for milking cows

We use your production data (shipped milk yield and components), feed costs (obtained daily from your feed software) and Class III milk prices (monthly milk price factors collected from the USDA official platform and adjusted for your farm’s protein and fat levels).

The calculation is based on class III milk prices and does not factor in any additional premiums or bonuses. If your milk contract is not a Class III contract, please contact us to adjust this.


Filters - sources and definitions

  • Pens are based on the DHMS.

  • DIM groups are set up by default according to the herd size and are currently not customizable.

  • Reproduction stages are based on reproduction events from the DHMS and include: open (not inseminated or open), inseminated (bred), pregnant (confirmed pregnant), dry (dried-off).

  • Lactation numbers are sourced from the DHMS

  • The status segmentation is set up on request from the farm owner/herd manager. This includes dry cows, milking cows and youngstock.

  • Production groups are set up on request from the farm owner/herd manager. These include high production, low production, far-off and close-up.

  • Custom production groups are personalized for each farm and set up on request from the farm owner/herd manager.

If you have questions or want to change the settings for states, production groups or custom production groups, please contact us via email or the support chat. Note that this can only be done by the farm owner/herd manager.

Did this answer your question?