Harvest Watch Wine Data: Behind the Dashboard with Anna Muthig, Data Analyst at InnoVint
Anna Muthig came to data analysis through wine, not the other way around. She spent her first years in the industry at a small tasting room and production operation making fewer than 2,000 cases a year, learning bottling, harvest, and winemaking from the ground up. She went on to make Riesling in the Clare Valley in Australia, then moved into sparkling at a larger facility where she first got serious about trials, data collection, and analysis. This year, while completing an analytics degree, she connected with InnoVint’s CTO Dan about a project they had both been thinking about: making real-time harvest data available to the entire wine community.
That conversation became Harvest Watch, InnoVint’s free public dashboard showing aggregated, anonymized harvest progression data across wine regions. Lauren Heindel sat down with Anna to pull back the curtain on how the dashboard was built, what drove the decisions about what to include and what to leave out, and how winemakers are actually using it.
Most winemakers develop their read on harvest conditions the same way people have always done it: informal conversations, running into colleagues, word-of-mouth at the end of a long day. Those exchanges are real, but they are based on tiny sample sizes and no systematic way to know if what you are seeing is consistent with what everyone else is seeing. Harvest Watch is built on the premise that the data to answer those questions already exists, and making it visible serves the whole community.
In this episode, we cover:
How an informal conversation about Brix levels among winemakers sparked the idea for a regional harvest data dashboard
The specific use cases that shaped the dashboard design: validation for individual harvest decisions, planning for custom crush facilities, and timing for shared equipment and cellar staff
What the dashboard shows and how to filter it to your sub-region and varietal for the most relevant picture
Why AQI and fire tracking belong on the same dashboard as Brix and acidity
How the first version of the dashboard was three to four times longer than what launched, and why paring it down was the right call
The feedback that shaped the tool during its first season, and what is planned for 2026
The origin: making shared knowledge visible
The idea behind Harvest Watch came from a pattern InnoVint’s CTO Dan had been observing for a while. Winemakers are constantly exchanging harvest information with each other, but those exchanges happen informally: at a bar, at the grocery store, in passing conversations that depend entirely on who happens to be nearby. The information is helpful when you get it, but it is anecdotal and based on whoever you ran into that day.
The question Dan brought to Anna was deceptively simple: why does this information not already exist in an accessible, systematic form? InnoVint has been collecting aggregated production data from wineries across regions for years. That data could answer the questions winemakers are constantly asking each other, not based on one person’s experience but based on the full dataset.
Dan's question, and then later his question to me, was sort of how does this not already exist, how do you not already know? How can we make a system that answers this question? All these questions that are constant, that people have all the time.
Anna Muthig , Data Analyst
InnoVint
Anna brings a practitioner’s perspective to why that information matters. She describes how winemakers in Carneros track the appearance of small handpicking bins along the highway as an informal signal that the sparkling harvest is beginning and it is time to start washing macro bins and cleaning presses. Everyone knows the season unfolds in sequence, varietal by varietal and region by region, and having data that reflects that progression rather than just your own vineyard or your closest neighbors gives you a fundamentally better picture.
If you have not visited Harvest Watch yet, start with your region and your primary varietal and look at how this season's Brix progression compares to prior years. In a year running late, like 2023, seeing that context helps calibrate decisions about when to pick even when your own vineyard's numbers look unfamiliar. The historical comparison lines are one of the most immediately useful features for that kind of calibration.
What the dashboard shows and how to read it
The goal Anna describes for the dashboard is flexibility: the ability to tailor what you are looking at based on your specific situation rather than presenting a one-size-fits-all view. The information most relevant to a custom crush facility expecting grapes from across the region is different from what a small estate winery needs when it is trying to confirm whether its Brix readings are in line with neighboring vineyards.
The map view allows filtering by sub-region, showing which areas are tracking ahead or behind regionally. The Brix and acidity charts show the seasonal progression for the filtered varietal and region alongside prior years, so you can see not just where things stand today but how this season compares historically. The tons-per-week chart captures the rhythm of the harvest, including the dips that correspond to storms or heat events, which provides useful context for planning.
What people are looking for first and foremost is validation. Are the things that I'm seeing reasonable? Especially when you're younger or you're starting at a new estate or you've gotten control of a new vineyard and you don't really have the same history. The calls that you're making affect your wine and the quality and the characteristics for the next year, so knowing you're in line with the rest of the region is this kind of added level of assurance.
Anna Muthig , Data Analyst
InnoVint
For custom crush and shared-facility operations, the cross-regional visibility has a specific practical use. If you are managing a shared press or split cellar staffing across multiple clients, knowing when the region’s Pinot Noir is winding down and Cabernet is ramping up is directly relevant to your scheduling. The dashboard makes that regional picture visible without relying on informal intelligence gathering.
When using the dashboard to compare your current readings to prior years, look at the shape of the progression curve as much as the absolute numbers. A harvest running late by two weeks may still be tracking a normal ripening curve, just shifted in time. The curve comparison tells you whether you are looking at a delay or an anomaly, which affects how you interpret what you are seeing in the vineyard.
Why AQI and fire data belong alongside Brix
One of the decisions that distinguished Harvest Watch from a straightforward production data display was the addition of AQI tracking and fire information. When the Pickett Fire broke out during the 2025 harvest season, Anna added those data points quickly, and the reasoning behind the decision captures something essential about how harvest decisions actually get made.
Picking decisions are not made solely on Brix and acidity. In California, and increasingly across the western wine regions, harvest timing is shaped by air quality conditions, proximity to fire activity, and the practical question of whether it is safe for cellar and vineyard crews to work. Those considerations are real and immediate, and they belong on the same dashboard as the production data because winemakers are weighing them at the same time.
Sometimes the calls that you're making aren't just based on Brix and acid. It's based on is the AQI low enough to harvest today? Like do we need to be getting masks for people in the cellar? Do we need to stop operations? Those data points are being considered right alongside grape analysis and phenological ripeness.
Anna Muthig , Data Analyst
InnoVint
The integration of environmental data alongside production data is part of what makes the dashboard a harvest decision-making tool rather than just a reporting surface. Anna describes the dashboard’s design principle as distilling data down to what people are actually basing decisions on in the moment, not presenting everything available but presenting the things that matter most when the season is in progress.
During smoke events, AQI varies significantly across short distances and can change rapidly. Use the AQI data on Harvest Watch as a directional indicator alongside your local air quality monitoring, not as a substitute for it. For cellar decisions specifically, your local reading from a monitor near the facility is the most relevant number, but the regional context the dashboard provides helps when deciding whether conditions are likely to improve or worsen over the coming days.
Building the tool: what got cut and what is coming
The version of Harvest Watch that launched was significantly leaner than the version Anna originally built. Her initial draft was three to four times longer, incorporating data on seasonal progression, bud break, veraison timing, and additional weather metrics. The paring-down process was deliberate: the goal was to arrive at what she calls the seasonal critical decision-making points, the data that winemakers are most actively consulting during the harvest window, without overwhelming users who were coming to the dashboard for the first time.
The dashboard before it was released started out three or four times as long as it is now. We paired it down to what we determined was kind of like the seasonal critical decision-making points. We didn't want it to be overwhelming, especially as people were first coming to it and determining how they wanted to work it into their workflow. We have a lot of ideas and a lot of places to expand.
Anna Muthig , Data Analyst
InnoVint
User feedback during the 2025 season shaped the tool in real time. San Luis Obispo was added as a region mid-season in response to demand. Formatting details, including standardizing decimal precision across different data types, came in through the feedback button and were addressed quickly. The feedback mechanism is anonymous by design, which means users who want a response to a specific question should include their contact information or reach out directly.
The plan for 2026 is to expand the regional coverage, with Lodi and Walla Walla among the areas users have requested. The longer-term vision is for Harvest Watch to become a national harvest tracking resource for the wine community.
If you used Harvest Watch during the 2025 harvest and have feedback on what worked or what you would want to see added, the feedback button in the lower left corner of the dashboard is the direct line to the team building it. Specific requests, including which regions to add and which data points would change how you make decisions, are exactly what shapes the next version. If you want a response to a specific question, include your contact information, since anonymous submissions cannot be replied to directly.
Watch or listen to the full episode on YouTube | Spotify | Apple Podcasts.
For more conversations like this, subscribe to Expert Talks and join The Punchdown, InnoVint’s online community for wine professionals.
Frequently Asked Questions about Harvest Watch Wine Data
What is Harvest Watch for wineries? +
Harvest Watch is a free public dashboard from InnoVint that displays aggregated, anonymized harvest data from wineries across wine regions. It shows real-time Brix and acidity progression by varietal and sub-region, tons crushed per week, historical comparison data from prior vintages, current weather conditions, and AQI and fire tracking. The dashboard is designed to give winemakers, vineyard managers, and custom crush operators a regional picture of harvest progression that would otherwise only be available through informal networks and word-of-mouth.
How do winemakers use harvest Brix data by region? +
Regional Brix data helps winemakers validate their own observations against a broader baseline. When a winemaker is making picking decisions on a new vineyard or in an unfamiliar year, knowing whether their readings are in line with, ahead of, or behind the regional average provides useful context. For custom crush facilities expecting fruit from across a region, regional Brix progression helps anticipate when different varietals will come in, which aids in scheduling equipment, staffing, and cellar space. The historical comparison lines on the dashboard allow users to see how the current season compares to prior years, which is particularly useful in unusual vintages.
What data does InnoVint Harvest Watch show? +
Harvest Watch shows Brix and acidity averages by varietal and sub-region, filterable by AVA and grape type. It includes a map view showing ripeness across sub-regions, a tons-per-week chart showing the harvest rhythm including storm-related dips, historical data from prior vintages for seasonal comparison, current weather conditions, and AQI with fire location and direction information. The dashboard was designed around the data points winemakers are most actively consulting during the harvest window, with a focus on supporting in-the-moment decision-making rather than comprehensive reporting.
Why does AQI matter during wine harvest? +
In California and other western wine regions, harvest decisions are increasingly shaped by air quality conditions alongside traditional ripeness metrics. When AQI is elevated due to smoke from nearby fires, winery operations may need to slow or stop, crews may require protective equipment, and smoke taint risk in the vineyard becomes a factor in picking timing. Harvest Watch integrates AQI and fire tracking on the same dashboard as Brix and acidity data because these factors are being weighed simultaneously by winemakers during the harvest season, not independently.
How can I provide feedback on Harvest Watch? +
Feedback can be submitted through the button in the lower left corner of the Harvest Watch dashboard. Submissions are anonymous by default, so users who want a response to a specific question should include their contact information in the submission. Requests for new regions, additional data types, and formatting suggestions have all been addressed during the tool's first season, and user feedback will directly shape what is added for the 2026 version.