NEWHow a B2B SaaS went from 4 to 131 AI answers citing it
TechDevs
ContactBook a free call

Case study

One screen runs a 70+ location brand.

One screen that runs a D2C brand: orders, inventory, cash, and what to do next

Book your free strategy callRead the strategy
9 hrs/wkof founder reporting time eliminated
  • Reorder SKU BLK-M (7 days of cover left; supplier lead time is 12 days)
  • Pause COD for 3 pincodes driving 41% of this week’s RTO
  • GM-40 sells 2.1× better bundled; push the bundle on the PDP
Client
Built for a 70+ location retail brand, productized since
Engagement
D2C operations dashboard
Led by
Prasad Iyer

The outcome

Where it started. Where it is now.

MeasureBeforeAfter
Daily tool loginsFive: OMS, IMS, Tally and two ad platforms1 screen
Founder reporting timeDecisions waited for the monthly P&L9 hrs/wkeliminated
Return to origin (RTO)Bleeding cash, unseen22%reduction

The situation

Every number existed. No one saw them together.

A D2C operator’s morning: OMS in one tab, inventory system in another, accounting in Tally, ads in two more. Nobody sees the whole business, so decisions wait for the monthly P&L.

What we found

  1. 01

    The morning started with the order system in one tab, inventory in another, accounts in Tally and ads in two more.

  2. 02

    No one saw the whole business at once, so decisions waited for the monthly P&L.

  3. 03

    Returns from a handful of pincodes were quietly costing money, and nothing flagged them.

OMS · orders
Order #10482Shipped
Order #10481RTO
Order #10480Delivered
Inventory · IMS
SKU BLK-MIn stock
SKU GM-40In stock
SKU WHT-SIn stock
Tally · accounts
LedgerLast closed month
ReceivablesPending
ReturnsNot booked
Ads platform A
Revenue, platform attributed
Ads platform B
Revenue, platform attributed

Monthly P&L

Decisions wait for month endNo one sees the five tools together

Illustration of the problem · names and orders are placeholders

The strategy

Join the data, then say what to do next.

A dashboard that only shows numbers moves the problem from five tabs to one. The plan was to sync the four systems the business already ran on into one model of the business, reconcile it against the books rather than the ad manager, and put an advisor layer on top that turns the joined data into the next decision: what to reorder, what to discount and where returns are costing money.

  1. Starting point

    Every number existed.

  2. Sync 01 · Orders

    Stream orders, returns and RTO from the OMS

    • Orders
    • Returns
    • RTO by pincode

  3. Sync 02 · Inventory

    Live stock cover per SKU, per warehouse

    • Cover per SKU per warehouse
    • Stockout alerts
    • Dead-stock alerts

  4. Sync 03 · Accounts

    Margins reconciled against the books

    • Revenue and margins
    • Cash position
    • Reconciled to the books

  5. Layer 04 · Advisor

    An advisor that names the decision

    • Reorder calls
    • Discount calls
    • RTO hotspots flagged

  6. Outcome

    9 hrs/wk

    of founder reporting time eliminated

Execution

What we did, move by move.

Each move below is a piece of the strategy above, with a drawing of the work as it ran.

Sync 01 · Orders

Stream orders, returns and RTO from the OMS

Orders, returns and return-to-origin rates arrive continuously from the order management system, broken down far enough to see which pincodes return the most.

  • Orders
  • Returns
  • RTO by pincode
OMS sync · orders and returnsStreaming
#10505BLK-M · CODPIN-FReturned
#10504BLK-L · CODPIN-EDelivered
#10503NVY-L · PrepaidPIN-BRTO
#10502WHT-S · CODPIN-CDelivered
#10501GM-40 · CODPIN-BDelivered
#10500BLK-M · PrepaidPIN-BRTO

RTO by pincode

PIN-B6
PIN-D3
PIN-E3
PIN-A1
PIN-C1
PIN-F0
Repeat offender3 pincodes

Illustration of the sync · orders and pincodes are placeholders

Sync 02 · Inventory

Live stock cover per SKU, per warehouse

The inventory system feeds days of cover for every SKU in every warehouse, with alerts for stock about to run out and stock that has stopped moving.

  • Cover per SKU per warehouse
  • Stockout alerts
  • Dead-stock alerts
IMS sync · days of coverLive
SKUWH NorthWH SouthWH West
BLK-M7d
BLK-L24d
GM-4033d
WHT-S41d
NVY-L22d
GRY-XL96d
Under 10 daysHealthyHeavyNot moving
StockoutBLK-MWH North · 7 days of cover
StockoutGM-40WH South · 9 days of cover
Dead stockWHT-SWH West · not moving
Dead stockGRY-XLWH North · not moving

Illustration of the sync · SKUs, warehouses and days are placeholders

Sync 03 · Accounts

Margins reconciled against the books

Revenue, margins and cash are reconciled against the accounting system, not the ad manager, so the number on the screen is the number the accountant will agree with.

  • Revenue and margins
  • Cash position
  • Reconciled to the books
Accounting sync · revenue this monthReconciling

Ad manager says

Platform A, attributed₹9.8L
Platform B, attributed₹8.6L
Revenue₹18.4L

Books say

Net sales booked₹15.1L
Less returns and RTO−₹1.9L
Revenue₹13.2L
  1. 1Orders both platforms claimed, counted once
  2. 2Returns and RTO netted out
  3. 3Matched line by line to the Tally ledger
Dashboard shows₹13.2LSource: the books

Illustration of the reconcile · all figures are placeholders

Layer 04 · Advisor

An advisor that names the decision

The advisor layer reads across all three syncs and writes the decisions hiding in the data in plain language. Flagging repeat-offender pincodes for returns is what cut RTO by 22%.

  • Reorder calls
  • Discount calls
  • RTO hotspots flagged
Advisor · next decisionsRead across 3 syncs
Reorder callIMSOMS

Reorder SKU BLK-M (7 days of cover left; supplier lead time is 12 days)

ReorderLater
RTO hotspotOMSAccounts

Pause COD for 3 pincodes driving 41% of this week’s RTO

Pause CODLater
Discount callOMSAccounts

GM-40 sells 2.1× better bundled; push the bundle on the PDP

Push bundleLater

How it ran

The engagement, in order.

  1. 1

    Step 1

    List the morning tabs

    Every system checked daily, and the question each one answered.

  2. 2

    Step 2

    Sync the sources

    OMS, inventory and accounting joined into one model of the business.

  3. 3

    Step 3

    Add the advisor

    Plain-language next actions written from the joined data.

  4. 4

    Step 4

    Productise it

    Built for a 70+ location retail brand, productised since.

The evidence

The screen that replaced five logins.

The shape of the live view: today’s numbers on top, the trend in the middle and the next three decisions underneath.

yourbrand · operations

Today

SourcesOMSIMSAccounts Synced
Orders today412+8.2%
Net margin31.4%+1.1pt
RTO rate6.8%-2.3pt
Stock cover23 days4 SKUs low
Orders trendLast 12 periods
20508012345678910111274
Next decisionsAdvisor
  1. 01Reorder SKU BLK-M (7 days of cover left; supplier lead time is 12 days)
  2. 02Pause COD for 3 pincodes driving 41% of this week’s RTO
  3. 03GM-40 sells 2.1× better bundled; push the bundle on the PDP
of founder reporting time eliminated
9 hrs/wk
RTO reduction after flagging repeat-offender pincodes
22%
replaces five daily tool logins
1 screen

The shape of the live view · yourbrand is a placeholder

What carries over

Three things to take into your own business.

  1. 01

    Reconcile to the books

    If the dashboard and the accountant disagree, nobody trusts the dashboard. Pick the books.

  2. 02

    Numbers are not decisions

    The advisor layer is the part people use every day. A chart rarely tells you what to do.

  3. 03

    Small leaks hide in joined data

    The RTO problem only showed up when orders and returns were read together by pincode.

In their words

What the client says about it.

4.9Average client rating

If you are looking for a dedicated team that understands real growth, work with TechDevs.

AshtonOperations & Digital Growth · Xhale City · 70+ locations

The service behind this

Operations dashboards

See how we run this as a repeatable engagement, not a one-off.

View the service

Next case study

Ranked inside ChatGPT before their competitors knew it was a channel

B2B compliance SaaS · US

Read it

Who led it

Prasad runs this kind of work.

Prasad Iyer, Co-founder · CTO, AI & operations

Co-founder · CTO, AI & operations

Prasad Iyer

Runs engineering and the AI division. Builds the systems the marketing runs on: multi-agent content pipelines, outbound booking agents and operations dashboards.

Read more

What happens next

  1. Answer a few quick questions about your business. About two minutes.
  2. Pick a time that suits you for a free 30-minute call with a founder.
  3. We study your site before the call, so you leave with a plan, not a pitch.
Book your free strategy call

Free, no obligation. Your answers save as you go, so you can stop and pick up where you left off.