Tally on Mobile: Why "MIS Reports" Has Been Broken for Indian Owners (and What Finally Works)

It is Saturday, 11 AM. The owner of a 120-person mid-market business in Surat is at a customer's office to close a renewal. Halfway through the meeting the customer pushes back: "Last year we did ₹42 lakh of business with you - I want a 12% discount this year." The owner needs to know, right now, the actual revenue from this customer over the last 18 months, the GST treatment, the credit notes raised, and the average payment delay. He needs it to negotiate with confidence. He needs it on his phone, on a 4G connection, inside the next two minutes.

What he does today: he WhatsApps his accountant. The accountant is at home. The accountant tries to remember the Tally login on his laptop. By the time he is back at his desk, runs the customer ledger, exports it to Excel, and replies, the meeting is over. The owner agreed to the 12% because he could not check the numbers in time.

This is the everyday tax that bad mobile MIS imposes on Indian businesses. It does not show up in any cost line, but it shows up in deals signed without the right context, in collections calls made with stale ageing, in board questions answered with "let me get back to you tomorrow". For an owner running a real business, the cost of not having Tally on the phone is measured in lakhs per year, quietly.

Three patterns are common, in roughly the order of how often we see them in Indian SMBs.

One - the WhatsApp PDF. The owner asks the accountant for a report. The accountant opens Tally on a desktop, runs the report, exports to PDF or Excel, and sends it on WhatsApp. The owner opens the PDF on his phone, squints at the columns, gives up, and asks a follow-up question by voice note. By the time the next file arrives, the original question has shifted. This is by far the most common Indian mobile MIS workflow.

Two - Tally On Mobile. Tally Solutions ships a mobile app called Tally On Mobile that mirrors a set of fixed reports from your Tally company. It is a real app, it works, and it is free to use. Limitations show up in the shape of the questions it answers - if the report you want is in the standard pack, you are fine. If you want anything ad-hoc ("show me Gujarat customers over 60 days with outstanding above ₹2 lakh"), you cannot get it from the app. Tally On Mobile is a viewer, not a query engine.

Three - paid connector apps. A small ecosystem of Indian vendors sells mobile dashboard apps that read Tally and push pre-built dashboards to your phone. Faster than the WhatsApp PDF, more flexible than Tally On Mobile, but the same fundamental constraint - the dashboards are fixed at the time the consultant builds them. Every new question your team thinks of becomes a ticket. Indian mid-market customers tell us their dashboard apps land in year two as "useful for the standard reports, useless for everything else".

The common thread is that all three options were designed around the assumption that you know your questions in advance. Tally On Mobile picked a list of standard reports to mirror. Connector apps build dashboards before launch. The WhatsApp PDF assumes the accountant already knows the shape of the answer when she runs the report.

Real Indian businesses do not work that way. The owner's questions change with the situation in front of him. On Monday he wants outstanding by ageing. On Tuesday he wants the same data sliced by sales rep. On Wednesday he wants to know which of those overdue customers also have pending credit notes. On Thursday he wants the GST input credit impact if he cancels three of those invoices. Five working days, five different shapes. No mobile dashboard built last quarter answers all five. The accountant ends up running one-off queries on her laptop and sending screenshots, which puts us right back at workflow one.

The other quiet failure of the existing mobile stack is the drill-down problem. A PDF or a fixed dashboard shows you a number. If you want to know "why is this number what it is" - which vouchers, which invoices, which credit notes - you have to call the accountant. The owner becomes a number reader, not an investigator. That is exactly the opposite of what mobile MIS should enable.

The shape of the right answer is not a better app. It is a different model. Real mobile MIS for an Indian Tally user should let the owner do four things from his phone, in 30 seconds or less per question, on a 4G connection, without calling anyone:

One. Ask any business question in plain English. Not pick from a menu of pre-built reports.

Two. Get the answer as a clean table or number, formatted for a phone screen. Not a 12-column Excel export squeezed into a 6-inch display.

Three. Drill down to the underlying voucher or transaction with one tap. So when the customer says "no, that ₹1.4 lakh was a credit note", the owner can show him the actual voucher in the next 10 seconds.

Four. Carry context across questions. Ask "Gujarat customers over 60 days", then ask "of those, which ones are with Anshu's team", and have the second question inherit the filter from the first.

None of the existing mobile Tally options check all four boxes. The WhatsApp PDF fails on all four. Tally On Mobile checks two and a half (it has fixed reports formatted for phone, but no plain-English query and limited drill-down). Connector apps check three (good drill-down, phone-formatted dashboards, but no plain-English query). The fourth check mark - context-aware follow-up questions - is what AI actually delivers.

The model that finally works is an AI analytics layer that reads Tally directly and runs as a web app you open in any phone browser. KolossusAI for Tally users works exactly this way. The owner opens kolossusai.in on his phone, logs in, types a question, and gets an answer in seconds. There is no separate mobile app, no app store install, no waiting for IT to approve a new download.

The questions can be anything Tally has the answer to. "Customer X total revenue last 18 months" is fine. "Outstanding above ₹2 lakh from Gujarat" is fine. "Compare our top 5 product margins this quarter vs last quarter" is fine. The AI translates the plain-English question into the right query against your live Tally data, runs it, and returns the answer. Each row in the answer drills back to the source voucher in Tally with one tap.

The phone-browser approach has a specific advantage that app installs do not - it works on any device. The owner's phone, his wife's tablet during a Sunday review, the laptop in the car, the desktop at his accountant's office. All the same interface, all the same data, all the same drill-down. The accountant is not the gatekeeper anymore.

To make this concrete, here are the actual questions Indian mid-market owners ask us they want to answer from their phone, drawn from POC conversations over the last year. Almost all are answerable today by an AI layer on top of Tally Prime.

From the customer's office. "What is our total revenue with this customer last 24 months?" "What discount have we given them historically?" "What is their average payment delay?" "Are there any pending credit notes?"

From the factory floor. "How much raw material did we issue to line 3 this week vs last week?" "What is the production yield this shift compared to the plan?" "Which vendor's invoice is pending payment that might delay tomorrow's dispatch?"

From the car between meetings. "What is our cash position this week vs same week last month?" "Which customers paid in the last 48 hours?" "Are there any GST returns or TDS deposits due this week?"

At dinner with family. "Quick check - how much business did we close this month? Are we ahead or behind last month?" The kind of question every owner thinks about constantly but cannot answer without firing up a laptop. With AI on phone, it takes 15 seconds.

Some buyers ask why we do not ship a native iOS or Android app. The honest answer is that for Indian mid-market businesses running Tally, a separate native app does more harm than good.

A native app means an extra install your owner has to approve. It means another login to remember. It means version updates that stop working when the phone OS changes. It means an app store review process every time we ship a feature. It means owners on older Android devices (still common in Tier 2 and Tier 3 cities) get left behind. None of this serves the owner trying to check a customer's payment history during a Saturday meeting.

The web-app approach delivers the same outcome - fast answers on phone - without any of those costs. The owner opens the URL, logs in, asks his question. iOS, Android, Windows phone, foldable, whatever. The same interface, the same data, the same drill-down. Updates happen on the server side. New features land instantly. The phone screen real estate is used the same way a native app would use it.

Two limits worth flagging upfront. First, AI on phone needs a working data connection for new questions. Once a question is answered the result is cached in the browser tab, but new questions require the AI to run on a server. Most Indian owners are fine with this because they already need data for everything else they do on the phone, but it is worth knowing.

Second, AI on phone is not a replacement for the desktop experience for heavy analytical work. If your accountant wants to do a full quarter-end variance analysis, she will still want a 27-inch monitor and a keyboard. The phone interface is tuned for the question-and-answer flow, not for deep multi-table investigation. Both work; they are tuned for different jobs.

On cost: a typical Indian mid-market deployment with one Tally Prime company and 5 to 15 users (mix of desktop and phone) lands at ₹2.5 lakh to ₹6 lakh per year all-in. The 14-day production POC is free, no credit card. There is no per-query meter, no separate "mobile add-on" fee, no per-device licensing. See Pricing for how the flat quote is shaped.

On setup: three weeks from kickoff. Day 1 to 3 we connect securely to your Tally and validate the numbers. Day 4 to 7 your finance team uses it from desktop and phone, and we tune phrasing for your business vocabulary. From day 8 the owner starts using it from his phone in real customer calls. The Saturday morning negotiation in Surat ends differently three weeks from now.